Category: Forecasting

Forecasting in Pharmaceutical Industry Using Artificial Intelligence: Current and Future Aspects

Abstract:

“Forecasting” The term broadly refers to the process of prediction as per the customer’s demand based on the huge historical sales data in the pharmaceutics industry. The aim of forecasting help to understand the market value and enable to predict the optimum level of customer demands. Thereby business management facilitates augmenting the future requirements from the previous sales quantity documents by considering both major and minor factors in a broad spectrum. This full-length Paper discusses the details of marketing, new product launch and specialized aspects such as orphans and bio-similar drugs. Artificial intelligence(AI) plays a strategic role to forecast the probable market requirements in advance for the industry and prepares to face future challenges. Forecasting could be multi-directional, application based on various approaches of the pharmaceutical industry such as Artificial neural network topology (ANN), Adaptive Network Based Fuzzy Inference System (ANFIS) which can be applied as a neuro fuzzy approach and proposed model approaches. This paper presents a detailed account on the key role of AI pertaining to the techniques that help the pharmaceutical industry supported by applications, illustrates, effectiveness and approach.

Keywords: Artificial intelligence, Artificial neural network topology, business management, forecasting techniques, pharmaceutical industry.

Introduction

“Forecasting” is the process of using the pattern contained in huge historical past sales data to predict future values. Forecasts predict the future levels of sales, demand, inventories costs, imports, exports, and prices among others in the form of numeric. The aim of forecasting is to guide the management to plan requirements for the marketing effort, material, personnel, production and market shares of the competitive products and marketing conditions are assisted. Clear and well-prepared forecasts should be accurate enough to allow for better future planning and control could not be validated without the forecast. Demand forecasting is one of the main inputs when developing long-term strategic plans. It is a method of analyzing the past and current historical data to determine future values. Hence, forecasting is the making of predictions about the future performance based on past and current huge data. Forecasting is necessary because nowadays, health and treatment services are facing issues in the conditions of prosperity and democracy. Therefore the pharmaceutical industry can be assist as a strategic sector.

1.1. Forecasting in the Pharmaceutical Industry

Forecasting are examined by the largest pharma market globally in the treatments for diseases such as rheumatoid arthritis and osteoarthritis, osteoporosis, carpal tunnel syndrome, tendonitis, rotator cuff tear etc. The segment accounted for 15% of the global total in 2017. Cardiovascular, oncology and anti-infective are drugs are the 2nd, 3rd and 4th largest markets.

Upto 2021, the fastest-growing segment of the global pharma market will be drugs for treating metabolic diseases, thyroid diseases and pituitary gland. This segment will grow at 10% a year going forward, the recent growth of 11.5% , but it will in 5th position for market size. The largest sub-segment of the global pharmaceutical industry are the anti-diabetic drugs worth over $85 billion in 2017; 2nd are the anti-virals and 3rd comes to the anti-hypertensives. Drugs for some of the less prevalent cancers like thyroid, skin, and ovarian cancer are the fastest-growing subsegments. The USFDA has allowed a less rigorous regulatory procedure and lower endpoint benchmark for cancer drugs, so increasing the rate of innovation.

2. Materials and methods

2.1. Structure of the pharmaceutical industry

It is one of the main processes of planning of forecasting in the pharmaceutical industry. It gives information about which products are purchased, when, where and in what quantities. By incomplete forecasting techniques, pharmaceutical manufacturers are affected. In a developed pharmaceutical market are seen on both valuable information data and market balanced power about products. In other sectors, obtaining the forecasting techniques’ by using terms & conditions, systematically sharing all available information and independently developing demanding scenarios from political terms and conditions with the greatest accuracy.

Fig2.1. Complex structure of pharmaceutical market

All pharmaceutical companies are in a close relationship with pharmacies (wholesalers & retailers), doctors, the pharmaceutical market, and patients. All pharmaceutical formulations should manufacture according to guidelines such as Food and Drugs Administration (FDA) and Goods Manufacturing Process (GMP). For the new challenges of the modern economy, pharmaceutical companies make a lot help to changes. For the new drug development manufacturing management and supply chain cause the effect. Pharmaceuticals companies will help the economy growth for many countries in the world. The population of the country and cure from a variety of medical conditions and benefits has helped the pharmaceutical industry an area of discussion.

2.2.The proposed methodology

For improvements of methodologies exist but in pharmacy straightforward, implicit and explicit assumptions of the direct human judgments and limited quantitative data. Two most common methods are the “Assumption Method” and “Morbidity Method”. It is based on the study such as historical data of past assumptions to predict future needs in human demand forecasting. In this study, a neuro-fuzzy approach was used for determining of next periods of the forecast.

2.3. Neuro-fuzzy approach

This methods is mostly used in management science. Accurately predicting sales forecasting performance in useful in many contexts in pharmaceutical sector management. When pharmaceuticals marketing managers review historical sales data accurate predictions help them to distinguish between suitable and unsuitable forecasting for pharmaceuticals products. The failure to perform an accurate forecasting demand decision may result in an unsuitable forecasting demand being approached in the pharmaceutical industry. Since the quality of the pharmaceutical industry is mainly reflected in its research and training, the approaches to forecasting demand affect the quality level of the pharmaceutical market. Accurate predictions enable pharmaceutical marketing managers to improve forecasting sales performance by offering additional customer supports such as giving some extra offers on purchasing. Thus, accurate prediction of customers is one way to enhance quality and provide better sales services. As a result, the ability to predict customers’ purchasing performance is important for sales forecasting in the pharmaceutical industry.

Neuro-fuzzy and neural networks are set of theory, which help to soft computing techniques or tools of establishing intelligent systems. A fuzzy inference system (FIS) employing neuro-fuzzy if methods are containing adequate knowledge from human experts can deal with imprecise and vague problems. FIS widely used in the optimization of sales forecasting, control and system identification forecasting. Fuzzy systems do not usually learn and adjust themelves.

2.3.1. Artificial neural network (ANN) topology

The artificial neural network (ANN) model is based on the biological nervous system,such as the brain working as information processing. They are loosely modeled after the neuronal structure of the mammalian cerebral cortex but on much smaller scales.

In other terms it is a simple mathematical model of the brain which is used to process the non-linear relationships between inputs and outputs in parallel like a human brain does every second. Artificial Neural Networks are used in a variety of tasks, a popular use is for classification. You can collects forecasting sales data for example of different years of sales data of pharmaceutical products and then train a neural network on the sales, then if you supply a new sales data of a pharmaceutical products it will give the new sales matches the models and then will output what sales of pharmaceutical products is.

Fig2.3.1. Artificial Neural Networks (ANN)

3. Forecasting techniques

There are five main types of sales forecasting techniques for pharmaceuticals. These techniques are; Simple conjoint type models, Zipf’s law, Simple Elasticity Model, The Bass Model, and Simple Extrapolation.

3.1. Simple co-joint type model

This model is based on working of how innovative your products is relative to the competition on the product quality and performance. Extensive historical data research has shown that the very simplest conjoint models in forecasting sales in the pharmaceutical industry.

3.2. Zipf’s law

In 1940, an eccentric Harvard linguist was developed a law that is known as zipping’s law. According to that law it can predict unconnected things such as the size of the market in a city, the population of the city, number of hospitals, number of pharmacies and also pharma market shares that’s all things are helped in the development of pharmaceuticals marketing or sale.

3.3. Simple elasticity model

If you any change your marketing investment i.e. 1%, how much will your markets share change? The answer gets this question is known as the marketing spend elasticity.

Did you know? The costs of pharmaceutical marketing behind brand drugs has a bigger impact on its sales forecasting than the costs of pharmaceutical marketing a generic class of medicines has on its sales.

3.4. The Bass Model

This model predicts the speed of uptake for a new brand drugs or generic class drugs in pharmaceuticals marketing. It is based upon the fact that when new products are launched in a market. People are nervous about new products they tend to react what other people are doing. But when they will accept the product without thinking of how other people get on first.

3.5. Simple Extrapolation

It is based on the fact when no new competitors are about to launch, no major clinical trial results data to be reported and there is no more accurate way to forecast than to just based on the current and future aspects. It is also based on established pharmaceutical products in a stable market. It is simple extrapolation method for forecasting of sales.

4. Current and future aspects of forecasting

Forecasting are influences many other functional areas within an organization. These linkages may be unidirectional or bidirectional. The links reflect the varied uses to which a forecast can be applied – such as revenue planning, production planning, resource allocation, project prioritization, partnering decisions, compensation plans, lobbying efforts, and so forth. These varied uses, and the effect of forecasting on many functional areas in an organization, reflect the first major challenge of forecasting – meeting the needs of varied and diverse stakeholders.

The link between forecasting sales revenue and unit volumes is an obvious one; however, the form of the forecast required may differ between these two. The unit volume forecast needs to include detail above that are required for forecasting demand – including information on sampling data, safety stock, and the distribution of products amongst the various packaging forms. These links should be bidirectional. In other words, the forecaster must understand the needs of the recipients of the forecast in order to select the best methods for generating the forecast.

In the most role of forecasting demand is to help in the development of business management in the pharmaceutical industry.

5. Conclusion

Forecasting should predict the number of product sales for each period of time in the pharmaceutical company. The case study given in the paper is based on a real-life example within the pharmaceutical field. The Paper presents review details on current and future aspects of forecasting in the pharmaceutical industry. And also discussed various approaches of the pharmaceutical industry such as Artificial neural network topology (ANN), Adaptive Network Based Fuzzy Inference System (ANFIS) applied as a neuro fuzzy approach, and proposed model approaches of forecasting in the pharmaceutical industry. Also, we discussed various factors affecting the parameters of forecasting in the pharmaceutical industry.

6. References

1. Journal of Military and Information Science Corresponding Author: Harun Reşit Yazgan ,Vol. 2, No. 2
2. Haykin, S., Neural networks; a comprehensive foundation, MacMillan College Publishing, 1, New York. (1994)
3. Jang, J.S. (1993), ANFIS: adaptive-network-based fuzzy inference system, IEEE Trans. On System, Man and Cybernetics. 23, 3, 665-685.
4. Forecasting for the pharmaceutical industry : models for new product and in-market forecasting and how to use them / by Arthur G. Cook. — [Second edition].
5. Preparing the Supply Chain Pharma Needs. (2014) “A.T.Kearney Pharma Supply Chain Panel 2014“. A.T. Kearney, Inc. Available: https://www.atkearney.com/web/the-purchasing-chessboard/article/-/asset_publisher/9AutfSQfJm6Y/content/preparing-the-supply-chainpharma-needs/20152. [Accessed April 30, 2018].
6. Erkollar, A., Goztepe, K., & Sahin, N. (2013). A Study on Innovation Performance Forecasting in Advanced Military Education Using Neuro-Fuzzy Networks, International Journal of Science and Advanced Technology, 3(4), 5-12.

Need of Sales Forecasting within the Company: Analytical Essay

Forecasting is nothing but predictions and hence closely related with the information flow of a supply chain. In this chapter, we will discuss the information flow in a supply chain from different perspectives like information sharing with channel partners, different obstacles in sharing information and bullwhip effect. Describing and analyzing the forecasting process is our purpose. A deeper understanding of the concept of forecasting will be achieved by discussing it together with the role of forecasting in various managerial functions. This will help us to understand how forecasting is correlated to these functions and the place of forecasting function in overall management. Next to this we will discuss both process of forecasting in general and different approaches of sales forecasting. Being a managerial function, we will discuss effective management of forecasting by seven principles which represents opportunities for management to improve the existing process. In the modern world, forecasting is becoming more dependent on computer technology and customized software. Multiple forecasting system (MFS) is a new trend in forecasting system and hence discussed with description of such software such as ‘Multicaster’. Further the forecasting method / technique changes are based on type of product. One could see different methods are available for forecasting. We have discussed a few of them which are relative to product range of companies studied. Since sales forecasting is management function, the performance measurement and finding out errors becomes an integral part of the process and is important for improvement in existing process. Hence we have discussed different measures of performance measurement at the end of the literature review.

Ordering focused forecasting

Traditionally forecasting was done based on orders from retailers. But these orders do not indicate real customer demand and variation in customer demand may increase during backward ordering. Each layer considers these orders as a base of their own demand forecast and pass the same to its supplier downward. A minute change in this demand may elevate inventory because this demand ordering moves back and each stage want to keep some safety inventory with them. In case of the random changes in demand, retailer may interpret it in wrong way and this could cause elevation or decrease of buffers in whole supply chain. As suppliers do not have direct visibility of this demand pattern they may never take their own decisions and hence rely on interpreted information from retailer. This elevates Bullwhip effect

Bullwhip effect: Distortion of information in supply chain

Bullwhip effect is the phenomenon where demand variability increases in supply chain as you move away from retailer to manufacturer. Lee, Padmanabhan and Whang (2004) had defined bullwhip effect at the first time as “the amplification of Demand variability from a downstream site to an upstream site.” It is observed that the order pattern in upstream direction is highly variable than downstream. The amplification in order variation may cause irrational decision making. (Lee et al 1997). See figure 2.1. It indicates the variation in demand between different stages in supply chain while moving upside. These variations are called as ‘Bullwhip Effect’. They also called as ‘Whip-Saw’ or ‘Whip-lash’. It causes dramatic effect on firm resulting in excessive inventory, poor product forecasts, insufficient or excessive capacities, poor customer service due to unavailable products or long backlogs, uncertain production planning and high cost correction like high shipments.(Lee et al 1997). There is no bullwhip effect in an ideal supply chain where as it is assumed that supply is stationary, fixed lead time, constant purchase cost overtime, no fixed cost of ordering and forecasting is not based on past demand ( Lee et al 1997).

Figure 2.1: variation in order pattern in bullwhip effect.

Forecasting Concept

Forecasts are nothing but predictions about future. May be forecasts of sunrise and sunset can be predictable without any mistake but it is not the scenario in business. Business equations changes as timegoes and hence prediction may give error. Mentzer and Moon (2005), describes sales forecast ‘as a projection into future of expected demand, given a started set of environmental conditions.’ One should not confuse the planning process and forecasting process. Planning is nothing but managerial actions which should be taken to meet or exceed the sales forecast (Mentzer& Moon, 2005). The aim of right forecast is to predict demand perfectly. Hence forecasting is necessary to be focused towards maximum accuracy. Where as planning is needed to be aimed towards efficacy and efficiency of all managerial functions to meet forecasting. In business each project starts with planning. But to plan, the prediction about future is needed so that one could prepare plan well in advance. Here the forecast comes in the picture. Forecasts have been used in all kind of companies, service sectors, and government organizations. Forecasts have been used as input to the planning project or set of activities. Hence Mentzer and Moon, ( 2005 ) say that forecasting is the focal point of corporate hierarchy.

A Summary of the characteristics of sales forecast are as follows:

1. Forecasts are always wrong and hence one should always expect evaluation of errors in it.
2. Long-term forecast are normally less accurate than short time forecasts. This is because larger standard deviation of error relative to meanshort-termt term forecasts.
3. Aggregate forecasts are normally more accurate than disaggregate forecasts. Aggregate forecast contains smaller standard deviation of error than disaggregate forecasts.
4. The Greater the distortions of information in supply chain the higher are the errors in sales forecast.

Need of sales forecasting within the company

Sales forecasting and Planning

Trade industries graft on the principle of satisfying customer demand through appropriate supply. Companies consider sales forecasting as an essential portion of this procedure. Figure 2.2 shows a simplified co relation

Figure 2.2: Sales forecast in operations planning and sales planning

End customers create demand, and practices such as promotions will increase this. Marketing therefore focuses on the formation of demand by end customers. Different strategies like serving other parties in this streamline like wholesalers and retailers make the same thing easier for the sales department. Supply should be sufficient to satisfy demand. Multiple management roles, such as production, procurement and distribution, work together to sustain inventory. In this chain, numerous distributors also play an important role. Constant flow of information flows through the complex structure of the various management functions and the parties involved. The process begins with demand and concludes with the functions of supply. Sales and operating system (S&OP) handles this information flow. For different companies, the S&OP process is different and can change as the environment changes. Lappide and Larry (2002) describe the operational control of S&OP. As we have already addressed this planning process, revenue forecasting acts as the basic seeding for the S&OP system. The forecast may be based on the study of past history of demand. The arrow of sales forecasting emerges from the demand side as the advertising role originates and handles the demand towards the end customer. Based on the sales forecast, the capacity plan is prepared by the supply side. The capacity plan is nothing but the ability of maximum possible inputs to satisfy demand. In order to consider approaches, both forecasting and capability plans are analyzed through the information network. In this process, Mentzer and Moon (2005) describe two major plans. Plan of service and plan of request respectively. Considering various time-to-time information gathered and approaches pursued, the request plans are provided from the S&O process. The demand plans provide sales and procurement divisions with an idea of future product release and action to achieve corporate strategies. Based on the available information; the S&OP operating plan is issued to provide functions. Each project is made up of various operational projects. Smooth S&OP operation requires precise forecasting. Continuing with the S&OP model of Mentzer, Armstrong (1983), demonstrates how the method of forecasting interacts with structured planning. See figure 2.3

Figure 2.3: Correlation framework for formal planning and forecasting

Planning is a series of business activities. Planning is to agree on priorities and to take appropriate action. The four planning steps: 1) identify goals; 2) develop strategies; 3) review strategies and take appropriate action; 4) track outcomes. Commitment to the fundamental goal is a key to success. But the exact prediction also plays an important role in the effective preparation and achievement of the final objective. One needs to understand that the difference between planning and forecasting. Forecasting is the assessment process of giving estimates and the planning process is the preparation of strategies based on these predictions. Double arrows indicate two-way flows of information between the database and the planning module, according to figure 2.3. An S&OP model operational plan was used to take action. Data can be sent for future and present use to the main database. Choosing the prediction method is an intermediate step between this task. The choice method depends on the company’s different needs. Figure 2.4 can describe it.

Figure 2.4: Use of forecasting methods for the various needs in the company’s planning

Sales forecasting and sales

Sales and forecasting are both closely interdependent as a managerial function. According to Miller and Heiman (1985), the forecasting function during sales planning should not be underestimated. The division of sales consists of different levels of management and sales force. Planning takes place on the basis of forecasting, and sales goals are defined region, time or service. Territory bases assist in the preparation of horizons and levels. That is, I.e. One can predict the product wise (SKU) or the product and place wise (SKUL). Time scales could be calculated by the existence of sales force commissions during forecasting.

Sales forecasting and finance/account

During corporate planning, the finance department carries important functions and plays a crucial role. The finance department determines on the financial level of spending on various consumer projects based on the forecast. This planning normally takes place annually. In some cases, preparation can last up to 5 years, such as launching new products and achieving long-term goals. In addition, the finance department will also manage the forecast-based corporate profit. Can division can target annual profit rates and, depending on policies, can last up to five years. Intervals are determined on the basis of product character and other terms such as once a month. (Mentzer & Moon, 2005)

Sales forecasting and production/purchasing

Output and average scheduling are closely linked to forecasting demand. Long and short-term forecasting is widely used in development and planning, according to Mentzer and Moon (2005). The long-term prediction should take place once planning is considered. When planning the selected product / product range for production; related functions such as selecting the right supplier, developing supplier relationships and planning the manufacturing plant’s cost structure are important. Building the whole system can take many years and therefore long-term forecasting is necessary. Plans rely on future sales of goods to be produced and site-wise sold.

The production plans rely on the purchase forecast during short-term forecasting. Wisner and Stanley (1994) suggest a close relationship between forecasting and purchasing. This shows the value of forecasting during the master purchasing plan preparation process. Purchasing activity requires time lags due to suppliers ‘ delivery and logistical intervention and therefore the purchasing department needs to know the scheduling outlook so that no stockout can take place. It helps to smooth out – of-stock production. (Mentzer & Moon, 2005)

Sales forecasting and logistics

The transportation department shall be held responsible for both storage and distribution from the storage site of produced goods to the destination. The logistics department, therefore, needs demand forecasting with SKU and SKUL rates. In preparation, both short and long-term forecasts are required. Long-term SKU-level planning is needed to decide warehouse and service storage capacity together. It is also necessary to consider transportation services during this long-term planning. The logistics department plans its own service plan based on the production plan and therefore forecasting plays an important role in the scheduling of logistics. The short-term forecast falls in the picture on an urgent basis and for tiny SKUsThis routine varies from daily (in some severe conditions) to order-based weeks or months. Companies usually either buy specialized logistics facilities from logistics providers from third parties or prepare them on their own. When buying or renting these services, the client should be aware of the service characteristics required. From the production forecast, this can be known. SKU and SKUL-based forecast are needed for this reason. (Mentzer and Moon, 2005)

Sales forecasting need in Marketing

The success of marketing is based on the company’s ability to meet customer demand and needs. Conditions such as stock outs and low innovativeness may lower demand and may result in loss of sales. A company is planning its activities in view of this principle. Refer to Figure 2.5

Figure 2.5: Sales forecasting need in Marketing: a flow.

Marketing plans are based on current demand, demand derived, pricing of competitors and various promotions. Knowledge of the forecast is needed to yield from the marketing plan. Annual rates can normally be considered and cycles can be either monthly or quarterly depending on the product (Mentzer & Moon, 2005; Armstrong & Brodie, 1999) Table 2.1 outlines the need for revenue forecasting in different administrative functions.

Table 2.1: Forecasting Requirements of Various management functions.

Forecasting in Supply Chain and Demand Management: Review of Literature

Review of Literature

This section explores selected journal articles in relation to the research objectives outlined in the introductory part of the research paper. The articles are critically reviewed to determine the key components and factors in forecasting, the forecasting techniques and models discussed, proposed and applied and the challenges of implementation in warehousing forecasting demand management. Furthermore, the research approach and methodology used by the studies will be evaluated to assess the assimilation of information and knowledge development.

Forecasting in Supply chain and Demand Management

Supply chain management (SCM) has become a critical concept in business management and consists of activities such as consists of sourcing, materials management, manufacturing support, and distribution management (Albarune and Habib, 2015) and revolves around managing material and information flow (Dugic and Zaulich, 2011). Supply chain management must anticipate the customer demand with products that in real sense have special character of qualities that fit in relation to their future area of use as well as being available at the time when the customer needs it (Dugic and Zaulich, 2011).

One of the critical strategies in customer demand management is forecasting; the concept defined as the prediction or estimation of an actual value in a future time period or for another situation (Albarune and Habib, 2015). Forecasting is an important determinant of operational performance (Acar and Gardner Jr, 2012) and is far most at the beginning of activities of SCM which initiates the all other actions.

Forecasting impacts and benefits are spread over across various departments and industries with the warehousing and distribution industry using forecasting as a planning tool (Albarune and Habib, 2015) and in rational decision-making based on numerical forecasted values. Its benefits are also diverse within Albarune and Habib (2015) asserting its impact in fulfilment of the customer requirements with some models giving better customer service than others (Acar and Gardner Jr, 2012), reducing risk and in improvement of supply chain process as well as inventory costs by increasing marginal improvements in average accuracy( ). Forecasts further add value by having place, possession and quantity utility (Dugic and Zaulich, 2011) and have daily impacts on different strategic, operational and tactical levels of a company’s forecasting needs (Fildes et al., 2019).

Optimal gain from use of forecasts depends on several key factors and element but mostly the choice of the forecast method and techniques with respect to the business model is the most crucial factor. In order to select the best model, the best practices, components and concepts should identify and aggregated in the best way possible. Briefly discussed below are the resource materials under critical review.

In the first journal article, Albarune and Habib (2015) examines the concept of forecasting in SCM and discusses the roles of Material Management (MM) and the marketing department team in forecast and proposes a forecasting management model that improves on spare parts forecasting and demand management based on Products’ character classification and demand pattern.

The second resource under review is the book chapter consisting of a cumulative of articles on various concepts and topics on business forecasting by the editors Gilliland et al. (2016) and provides general guidance on important considerations in the practice of business forecasting. The chapter covers the essential elements, uncertainty, measurements of forecastability and provides guidelines for improving forecast accuracy.

Third, a master’s thesis by Dugic and Zaulich (2011) conducts a case study on a furniture manufacturing and distribution company, IKEA, to identify the forecasting technique and model used and the findings to make an analysis of associated challenges and recommendations to a better forecasting performance. The manual and automatic system used by IKEA where planners demand and Sales Response System respectively forecast demand and strategically meet the demand requirements as well as make daily adjustments to increase forecasting accuracy.

Fourth, an extensive article by Fildes et al. (2019) explores the forecasting factors and problems faced by large-scale retailer alongside the on the strategic, tactical and operational level of forecasting needs as well as the market level, chain level, and store level aggregate sales forecasting based on production units, location, time buckets or promotion. Furthermore, different forecasting methods and benchmarks are discussed at product levels and a comparison on forecasting accuracy conducted

Lastly, Acar and Gardner Jr (2012) designs a forecasting model for a global manufacturer using a mixed-integer program that forecasts demand to give values for production, inventory, and transportation planning with the aim of minimizing total supply chain costs. The model takes tactical planning of (Fildes et al., 2019) at various levels of production, exponential smoothing methods explained in Dugic and Zaulich (2011), different initialization and fitting procedures and a trade-off analysis to evaluate the operational performance of the forecast model.

Key Components and Factors

Various components and factors that are imperative for the effective functional of warehousing and distribution firms and business with regard to the forecasting aspect has been identified in literature subject to their respective forecast and business models. For instance, (Corr, 2012) outlines seven elements namely the time frame for analysis, direction based on the baseline, magnitude in terms of distribution, probabilistic point or range value, range, statistical confidence levels, and historical forecast error for similar forecasts that every forecast should consider for it be valuable. Dugic and Zaulich (2011) also asserts that the most important attributes when selecting the best forecasting method as reliability, time, cost and the ability to identify market changes. Other key components and factors are discussed below.

Collaborative Forecasting.This concept concerned with departments working together to facilitate effective forecasting and is an important tool in meeting lumpy demand (Dugic and Zaulich, 2011). It reduces the risks from making inaccurate focuses by facilitating flow of adequate information in a timely and convenient manner. Schedules regarding when the meetings and information exchange should occur are created and support communication systems established (Fildes et al., 2019). Furthermore, structuring the organization and departments based on collaborative management principles such as value adding manager-subordinates relationship boosts the trust among staff and creates a productive workplace (Albarune and Habib, 2015).

Forecast Errors and Uncertainties.Forecasts cannot be utterly correct, and every forecast should include an estimate of error (Gilliland et al., 2016). Forecast errors exists when actual results differ from the projected value (Albarune and Habib, 2015), when forecasts are presented in the form of a single event or a single number (Gilliland et al., 2016), when actual demand is uncertain, and due to truncation of distribution values (Acar and Gardner Jr, 2012). Aaccording to (Boylan, 2003), forecast errors need to have an upper bound and lower bound error metrics for to generating exception reports and inform corrective actions in the process of forecastability. Proposed methods to reduce forecast errors uncertainties include improve statistical forecasting methods, revising judgmental statistical forecast and identifying more forecastable series (Dugic and Zaulich, 2011; Boylan, 2003), use of probabilistic forecasts tools such as prediction intervals, use of fan charts and probability density charts as better tools (Goodwin, 2014), sharing real time data and information across the SCM network using electronic communication technology as propagated by the CPFR model (Albarune and Habib, 2015),

Forecastability.The term is associated with the mitigation of forecast errors and is an important factor if the uncertainties and forecast accuracy is to be improved. The term may refer to the degree of accuracy when forecasting a time series, the smallest level of forecast error achievable (Corr, 2012) and the comparisons of forecast accuracy with a benchmark (Schubert, 2012). The forecastability values are essential in establishing comparability through benchmarking and use of metrics to size the forecast error (Dugic and Zaulich, 2011).

Demand Management.This is the anticipation and effort to approximate the customer demand for the product and using this information to make organization decisions with the aim of not only meeting the demands but also reducing demand for least profitable items and processing demand for the most profitable (Dugic and Zaulich, 2011). The ability to accurately forecast the demand is critical to the survival and growth of a retail chain because many operational decisions such as pricing, space allocation, availability, ordering and inventory management are directly related to its demand forecastt (Fildes et al., 2019).

According to Albarune and Habib (2015), the forecast managers have to ensure demand is met and should hence have alternative plans in case of any deviations in initial forecasts. In addition, sharing of knowledge about consumers, technology, logistics challenges and opportunities with other member in the supply chain allows timely response to customer demand without a compromise in customer service (Dugic and Zaulich, 2011). Also, it is essential that the business understand the difference between true and constrained demand as this influences the respective forecasts. True demand is largely unobservable and requires forecasting right at the beginning of the planning process (Fildes et al., 2019) while the constrained demand that is obtained by considering the limitations of goods and services provision in response to demand allows for forecasting in the light of these limitations (Gilliland, 2010).

Product Categorization. Accurate and effective demand management relies on customized forecasting where products are categorized into so that respective demand patterns (smooth, erratic, low turnover, slightly sporadic, and strongly sporadic) can be established (Albarune and Habib, 2015). Besides being used in demand management, product categorization and classification is also important for balancing between customer service and value creation, easing of forecasting process, keeping inventories at predetermined levels and to be able to make an analysis of the company’s assortment (Dugic and Zaulich, 2011).

Forecasting Techniques and Models

A variety of forecasting techniques are discussed through out literature. Dugic and Zaulich (2011) notes that the techniques are categorized into qualitative, extrinsic, and intrinsic methods while other authors group them into qualitative and quantitative techniques (Albarune and Habib, 2015). Qualitative techniques are mostly used for large product categories and base the customer demand on judgment and perception such as experts’ opinion, market research analysts and the Delphi method. On the other hand, extrinsic techniques base their forecast on external events and information affecting the demand and are suitable for product group or product families while the intrinsic technique use historical data to predict future values and mostly uses quantitative forecast models (Dugic and Zaulich, 2011).

Various models are also discussed, and others proposed in the journal articles and the book chapter under review.

Albarune and Habib (2015) proposes a forecasting management model that aims to overcome the loopholes in workflow due to behavioural user attitudes that bring distrust and poor coordination among departments. To overcome the limitation of forecasting practices identified by the authors, the model applies the concepts of product categorisation (Dugic and Zaulich, 2011), and has value addition and follow up in every steps of any activity for essence of quality control.

In discussing the concept of forecast ability, Gilliland et al. (2016) discusses various approaches to forecasting. One, forecasting software is using automatic model-selection procedures has been used to give an immediate lower bound for error reduction in addition to use as a benchmark for manual models (Boylan, 2003). Second, in finding a more forecastable series to forecast, Boylan (2003) also proposes the use of an autoregressive model of order one where current demand is related to the demand in the previous period and hierarchical models to address seasonality in demand. Fildes et al (2019) discusses three product hierarchies that can be used for planning forecasts namely SKU level, brand level, and category level for weekly, promotional planning and assortment forecasts respectively.

In the consideration of the principles and concept of demand management and its associated influential drivers of demand such as stock outs, intermittence, seasonality, calendar events, weather, promotions and social media reviews, Fildes et al. (2019) reviews different classes of models at product level demand subject to product hierarchies, namely univariate, multivariate and econometric methods. Since the uni variate methods do not consider external factors such as price changes and promotions as drivers of demand, they are only suitable for higher aggregation demand forecasting. It is therefore necessitous that base-times and judgmental adjustments (Boylan, 2003’s third strategy to better forecasts) are applied to lift effect of the most recent price reduction and/or promotion, and also the judgements made by the managers (Fildes et al., 2019) and the logistics departments (Dugic and Zaulich, 2011). The author also discusses econometric models mostly linear and dynamic regression models which take into account the mentioned factors of demand with advantages of the models being “simple, easy and fast to fit” (p 37).

Lastly Acar and Gardner Jr (2012) incorporates Boylan (2003)’s strategy that selecting the best forecast method is crucial for forecast ability and uses the trade-off analysis technique as well as the one-step errors from Morlidge (2013)’s Naive forecast method to get scaled error measures for different forecasting methods and hence determine the most suitable forecasting model.

Challenges

Various challenges have been identified in the above reviewed materials. Albarune and Habib, (2015)’s qualitative model has trust issues among users due to behavioural attitudes, difficulties due to use of revised forecast values in the freezing period, and lack of skills. Also, according to (Dugic and Zaulich, 2011), demand management is met by issues of lack of synchronization between departments, overemphasis on forecast demands and supply-demand misalignment making forecasting difficult. Benchmarking which plays a vital role in forecastability to minimize errors is also a challenge due to limited knowledge on the metrics (e.g. lead time and level of aggregation) used to model the benchmark and incomparability data from forecasts surveys due to differences in industry and product, granularity, forecast horizon, forecast process, and the business model (Kolassa, 2008).

References

1. Acar, Y. and Gardner Jr, E.S., 2012. Forecasting method selection in a global supply chain. International Journal of Forecasting, 28(4), pp.842-848.
2. Albarune, A.R.B. and Habib, M.M., 2015. A study of forecasting practices in supply chain management. International Journal of Supply Chain Management, 4(2), pp.55-61.
3. Boylan, J., 2009. Toward a more precise definition of forecastability. Foresight: the International Journal of Applied Forecasting, (13), pp.34-40.
4. Corr, C.K.R., 2012. What Demand Planners Can Learn from the Stock Market. The Journal of Business Forecasting, 31(3), p.21.
5. Dugic, M. and Zaulich, D., 2011. Forecasting system at IKEA Jönköping.
6. Fildes, R., Ma, S. and Kolassa, S., 2019. Retail forecasting: research and practice.
7. Gilliland, M., Tashman, L. and Sglavo, U., 2016. Business Forecasting: Practical Problems and Solutions. John Wiley & Sons.
8. Gilliland, M., 2010. Defining’ Demand’ for Demand Forecasting. Foresight: The International Journal of Applied Forecasting, (18), pp.4-8.
9. Goodwin, P., 2014. Getting real about uncertainty. Foresight: The International Journal of Applied Forecasting, (33), pp.4-7.
10. Kolassa, S., 2008. Can we obtain valid benchmarks from published surveys of forecast accuracy? Foresight, 11, pp.6-14.
11. Morlidge, S., 2013. How good is a’ good’ forecast? Forecast errors and their avoidability. Foresight: The International Journal of Applied Forecasting, (30), pp.5-11.
12. Schubert, S., 2012. Forecastability: A new method for benchmarking and driving improvement. Foresight: The International Journal of Applied Forecasting, (26), pp.7-15.

Ways to Make Sales Forecasting More Accurate: Argumentative Essay

Abstract

This paper explores seven published articles that talk about the importance of making an accurate Sales Forecasting. This paper will answer how Sales Forecasting helps business people or others because it tells the seller the future trends of what people in general like, and the sellers can know what their customers want. Therefore, the seller and company owner can create a bright and reasonable plan for sales. All the business people require to learn Sales Forecasting for establishing a company or sales goods. Mentzer (2005) and Moon (2005), Moskowitz (2007), Gofman (2007), Hyedima (2018) say that data is hard to collect, but Sales Forecasting can easily collect data and show the information to the company. Other authors provide Sales Forecasting background information. This paper continues researching the function of Sales Forecasting and how Sales Forecasting applies to daily life. Learning Sales Forecasting, no matter business students or other people, is a advantageous decision.

How to Make Sales Forecasting More Accurate

Currently, there is an increasing number of people in the business field who want to earn a lot of money. For business students, there are so many essential sections that they need to cover. Sales Forecasting is the most basic knowledge they need to know before they establish a company or study for business courses. The most significant point that people have to learn about Sales Forecasting is that Sales Forecasting can affect companies’ plans, budgets, and the money that they earn for each quarter. In order to become successful, business owners need to have the skills of Sales Forecasting.

The Sales Forecasting is estimating all the products or the number of the products’ money they can earn in a period. Although Sales Forecasting is crucial, to reach a high quality of Sales Forecasting is not easy. Before using Sales Forecasting, people should choose the most proper way to estimate. Business owners must learn about what Sales Forecasting effects and what factors that Sales Forecasting cause is essential. This paper will explicate two elements that create an accurate Sales Forecasting, two reasons why Sales Forecasting makes business earn more money and future trend in business, and how Sales Forecasting stops people from being overtaken by future changes, and also it provides value of being proactive.

Making Accurate Sales Forecasts

Why is it so crucial to make an accurate Sales Forecasting? Sales Forecasting works for any company, no matter the size of the company and the members of the company. Sales Forecasting is an estimation of how much money a company earns in a specific time or how many goods a company sells to their customers in a period. To make a Sales Forecast, people should learn how it works, consider all the factors that the company may face in the future. The function of Sales Forecasting is Sales Forecasting can encourage sales members and give them a goal to sell products, and companies can use Sales Forecasting to arrange how many products should be produced and prevent overstocked commodities. The several elements of how to make Sales Forecasting more accurate are outcome factors and internal factors.

In ancient time, people built a small village as a tribe with their friends and family members. Sometimes, they would be invaded by other tribespeople. People called it outcome factors. In the Sales Forecasting, people who want to make Sales Forecasting more accurate have to consider outcome factors. Demand trend, economic change, horizontal competition trend, and the government and customers spending trend are counted as outcome factors. The most critical element under the outcome factors is the demand trend. The demand trend is what is the mainstream of current society is, like fashion or politics, what the hobby of the people is, in general, like and mobile population in the contemporary community. Hyedima (2018) points out that learning what kind of hobbies of the people in general like is essential. The company provides a service that listens to what customers want and follows their ideas to create a product that fits them. To ensure this service, the company owners should learn what the leading fashion stream of current society is and present multiple suggestions to their customers. The company also needs to know about how many residents in every region, which helps companies to create a statistic graphic, where can they can build a store, and where has the largest population is. Therefore, company supposes to collect marketing information of their customers and make sure the quality of their service and products. The second element is the change of economic. Economic factors can affect whether customers will purchase goods or not. For improving the accuracy of Sales Forecasting, an enterprise needs to focus on its products on the market and supplier information such as the quality of their manufacturers and factors. With the development of current technology, online shopping becomes a popular way of how customers buy goods. For making an accurate Sales Forecasting, people should concentrate on resource distribution and future development. The third element is a horizontal competition trend. Estimating correctly one’s strength as well as that of one’s opponent is a phase that the Chinese elderly speak often. How many goods that a company can sell are affected by their competitors because customers have more opportunities to buy goods that more approaching what they want. Some people consider the appearance of a product, while others may think about the performance of the product. It is necessary to know what different types of people that the opponent company targets, what prices the opponent company sets, and what service they provide. The last outcome element is the government and customer spending trend. Moskowitz and Gofman (2007) say that it is vital to catch up on the information, both with the people and the government. Sometimes, the government posts a new policy, or customers tell companies what they need can be used to make an accurate Sales Forecasting.

In ancient time, people battled another tribe because they would like to expand their land. Usually, there is a king who rules a land. The king would be worried about when two or more than two tribes fighting; they may take away his position. Therefore, the king of the land needs to figure out this circumstance, and people named this situation as an internal factor. Mentzer and Moon (2004) mention that Sales Forecasting is to provide relevant data to all the companies, no matter which field they are in, to make sure those companies can find out the potential problems that they are confronting now. There are four elements under the internal factors that are a marketing strategy, sales policy, sales members, and production status. Creating a sales strategy can identify the position of the goods. For example, Burberry is a luxury brand in marketing. They set a high price, which is only for the rich, and H&M sets their prices lower than Burberry because they would like to sell their products to students or people who do not have a vast amount of money. Sales strategy needs to consider how much money their company wants to earn and what kind of people they would like to target. Sales policy is viewed as a way how a customer pays and refunds. A company should provide many payment methods to their customers. Some people do not bring a credit card when they go shopping; therefore, shoppers should provide second payment methods for their customers. Sales members are also crucial for a company that sells more goods. Humans are emotional animals. When buying products, they are always affected by the service that the company provides. Sales members should be trained and make themselves look professional. As a result, customers can decide to buy their goods immediately. The last factor is making sure that the company has enough stocks for selling. If a company wants to face one hundred customers, but only prepare fifty goods, it is not appropriate.

In conclusion, the two critical elements of how to make accurate Sales Forecasting are considered inside factors and outside factors. Knowing what people want, economic change, competitors’ advantages and disadvantages, government orders, company strategy, selling policy, sales members training, and enough stocks are essential elements for making an accurate Sales Forecast. Making a good Sales Forecast is not only working in the business field, but it also provides a method for the people in general to learn the importance of learning information ahead and decide to deal with the problem that people may face.

How Sales Forecasting Helps People Gain More Money

The previous section talks about the Sales Forecasting helps sellers gain more money by learning this skill. How can a seller earn more money from their customers? Nowadays, the quality of human life is increasing. Many people are able to buy their goods, which makes them have a sense of satisfaction. From a seller’s perspective, they are confused how to sell goods to their customers and make their customers have a sense of happiness and pay for their goods. Therefore, the sellers need to learn Sales Forecasting before they sell goods to their customers. The two reasons of Sales Forecasting can help people gain more money are sellers need to learn about their customers what they want and sellers can see the trends of how people in general like.

Customers would like to pay for goods that are exactly the same as their imagination. A customer always focuses on goods prices, goods performance and goods practicality. Some people think the outlook of the goods should be pretty, while other people may think the best performances is more important than its appearance. Shipley (2019) points out that wanting to make an accurate Sales Forecast, people need to figure out customers what they want and what goods are suitable for themselves. Customers will pay money for goods, when thinking the goods is worth its prices. Those examples can present a fact that people’s perspective is different. Moskowitz and Gofman (2007) say that the most effective way of improving company’s profits is learning what their customers want. Learning what customers need, company can give a better service for their customers. When customers satisfy of what they gain from company, a company can obtain an unexpected benefit. If the customers satisfy, a company can open the marketing better and easier. When company has a stable group of customers, customers will be going to introduce products to other people. This is the best way how to advertise a company’s goods and it is the cheapest way to advertise. On the other hand, humans are an emotional animal. Many times, humans make a decision with their emotion. If the customers like the services that the company provides, the customers are willing to give the company a hand, when they have the resources. Therefore, sellers are supposed to learn about their customers’ ideas individually. The Sales Forecasting can teach a method of learning customers’ mind.

Sellers can predict if the kinds of goods they are going to produce will appeal to people who buy them. In the business field, learning customers mind and making goods only for themselves is a correct method of earning money. However, an enterprise that wants to improve has to consider the people in general. According to Moskowitz and Gofman (2007), Sales Forecasting presents several methods of learning how sellers can learn to sell goods to more people and how sellers can predict future trends that customers would like to purchase. For example, Apple company always produces iPhones with different colors, storages, and function. The reasons why Apple does this is because they would like to cover all the people in general. Some people do not have enough money, so that those people may purchase a device with a lower storage. Some people do not like classic colors like black and white; therefore, they can buy green, yellow and blue as they like. In this way, the seller can earn money from learning general people’s ideas and produce a product corresponding to these people.

In conclusion, Sales Forecasting can help people gain more money by learning about what their customers want and about the trends that people in general like. Providing goods that customers want and producing goods that everyone can accept are the main ways that sellers can earn money from their customers. Learning Sales Forecasting and using this skill properly is a necessary skill to success.

How Sales Forecasting Works in The Marketing Planning Process

The earlier section mentioned that the people in business filed need to understand how the Sales Forecasting helps people to gain more money. Consumers can also gain enough money for themselves. Nowadays, many people are facing a big challenge that people who are in business will stagnate and be overtaken by the competition. There are increasing number of students who are graduating from the University, which means it will be difficult for them to find a job and earn money for themselves. Some people consider establishing a company or working with a mature company because they can earn a huge amount of money or a good salary. In this situation, Sales Forecasting shows its important meaning. The reasons of Sales Forecasting work in the marketing planning process are that Sales Forecasting can prevent people from being overtaken by changes in the future and it shows the value of being proactive.

Estimating correctly one’s own strength as well as that of one’s opponent will help people make decisions more carefully and cautiously. Modern businesses do not know how to create a new product and produce it. Most businesses are only changing one product’s appearance or improving it performances, but rarely people have noticed that they need to invent new products. It shows the importance of the Sales Forecasting. Mariotti and Glackin (2007) mention that Sales Forecasting is based on the full consideration of various factors in the future, which combined the sales performance of the enterprises, and it also uses a certain of analysis method to predict a Sales target. The ancient people usually went hunting small creatures for themselves as a family. This is the best way to gain food for survival. There were another group of ancient people who went hunting with more than two families because they could not kill a giant creature by a group. In that period, many people would think that each family would not obtain enough food for themselves. In fact, the second group of people gained more food than the previous family. This is similar to the business field. In order to survive, the company must master all the activities of the opponent companies or competitors in the marketing. For example, what kind of people is the competitors’ target market, and what prices the competitors set. The reason business people need to know this information ahead is because sales are heavily influenced by competitors. According to Hyeduma (2018), Sales Forecasting can provide an idea what inventors can make and give warning ahead for beneficial action. In 1999, nobody could imagine cell phone could be used as lighter, camera, and gaming system. The reason the people have touch screen cell phone is because of Steve Jobs. He was a successful businessman and inventor because he knew the Sales Forecasting, what kind of invention can fit people in the future, and what appeal more to people in order for them to purchase his inventions.

Being proactive is crucial for all the people no matter what their circumstances. Sometimes, people regret when they are doing something incorrectly. Therefore, knowing the information and its consequences will be necessary. Samsung’s profits fell to 56% in third quarter 2019 because of cell phone exploration scandal. Hyedima (2018) points out that people should be aware of the importance of accurate Sales Forecasting to different parts of the business field. Any pieces of problems can cause a huge problem. To prevent a company facing bankruptcy or scandal, businesses should use the knowledge of Sales Forecasting to learn about the information ahead. Therefore, businesses can gain a brighter future by learning about Sales Forecast.

To sum up, Sales Forecasting can prevent people from being overtaken by changing of the future development and provide more information for people to make a proper decision. Sales Forecasting can provide ideas to the manufacturers that products should be innovative and that inventors and companies should be careful when doing business. When learning Sales Forecasting, people can prevent a lot of problems that people usually do not focus on, and people who have had studied Sales Forecasting can make their company better.

Conclusion

In conclusion, this paper talks about how people make Sales Forecasting more accurate, how Sales Forecasting enacts in the marketing planning process, and how Sales Forecasting important to business people. This paper reports what outcome factors and internal factors are, the importance of learning what customers want and what future people in general like, and what the reasons for people will not be overtaken by learning Sales Forecasting are. The importance of making an accurate Sales Forecasting is by learning Sales Forecasting, and people can encourage their sales members to be positive, promote their sales members to sell more goods, and show the values of companies’ products. Enterprises can know the number of assets that they need to produce, which can prevent the goods overstock. Furthermore, companies can improve the quality of the products and make a proper schedule. Sales Forecasting can be seen as a system that inputs a piece of certain information when using this skill to convert to new information, which provides a company with the suggestion of what they should do. To know the target products, business people are going to sell or produce, collect, and to analyze information. This paper points out many advantages of learning Sales Forecasting and the consequences of not using Sales Forecasting. Therefore, people who learn Sales Forecasting can make their companies better. Future research on Sales Forecasting can investigate how Sales Forecasting can even better help both small and medium enterprises to become successful.

References

1. Hyedima, A. (2018). Sales Forecasting System. Retrieved from https://narr.ng/jspui/handle/123456789/56.
2. Mariotti, S., & Glackin, C. (2016). Entrepreneurship: starting and operating a small business. Boston: Pearson.
3. Mentzer, J. T., & Bienstock, C. C. (1998). Sales forecasting management: understanding the techniques, systems, and management of the sales forecasting process. Thousand Oaks: Sage Publications.
4. Mentzer, J. T., & Moon, M. A. (2005). Sales forecasting management: a demand management approach. Thousand Oaks, CA: Sage Publications.
5. Moskowitz, H. R., & Gofman, A. (2007). Selling blue elephants: how to make great products that people want before they even know they want them. Upper Saddle River, NJ: Wharton School Publishing.
6. Moon, M. A., Mentzer, J. T., & Smith, C. D. (2002, June 3). Conducting a sales forecasting audit. Retrieved from https://www.sciencedirect.com/science/article/pii/S0169207002000328.
7. Shipley, L. (2019). How to Make Your Sales Forecasts More Accurate. Retrieved from https://hbr.org/2019/08/how-to-make-your-sales-forecasts-more-accurate.

A Comparative Evaluation of Monthly Electricity: Consumption Forecasting in Sri Lanka

Abstract

Sri Lanka as a developing country, over 98% households have been electrified and it is crucial to plan for future electricity demand in order to match the demand with supply. This study aims at forecasting monthly electricity consumption in Sri Lanka and explores the weather influence on that electricity consumption. Due to higher living standards, weather has a considerable impact on the short-term electricity demand due to the use of fans, air conditioners and refrigerators. There are three main weather-related factors that affect demand they are Rainfall, Humidity and Temperature. In this study, four forecasting approaches (Classical Decomposition Method, Exponential Smoothing Method, Probabilistic Modeling in Time Series, Autoregressive Distributed Lag (ARDL) model Approach) were employed in forecasting monthly electricity consumption in Sri Lanka. Additionally, Artificial Neural Network (ANN) was used in order to investigate the applicability on forecasting monthly electricity in Sri Lanka as an advanced method. Arc-map software was used for spatial computations and twenty meteorology stations were considered to spatially interpolate the weather data by using the Inverse Distance Weighted (IDW) interpolation method. Under this study, it was revealed that ARDL and ANN model approaches which the weather influence was incorporated perform better in monthly electricity consumption forecasting.

Keywords: Electricity consumption forecasting, Weather impact, Inverse distance weighted interpolation, Autoregressive distributed lag, Artificial neural network.

Introduction

Electricity that is delivered to end-users like households and businesses is the result of a complicated operational process starting from Generation, Transmission, Distribution and real time supply-demand balancing. Ceylon Electricity Board (CEB) being the only generation and transmission utility in Sri Lanka is responsible for this total cycle and this operation is controlled centrally by the System Control Centre (SCC) which is the nerve center of operational activities of the CEB. As electrical power in Alternating Current form (or AC in short) cannot be stored in real-time, the SCC need to match the demand for electricity in the country at any given time with exact amount of generation in real-time. SCC needs to meticulously plan this real-time supply-demand balancing act and this type of planning is called “operations planning”. Operation planners need to plan for time steps from next week to several months ahead. One of the fundamental prerequisites of Operations Planning is short-term demand forecasting.

The main objective of this study is identifying the correlation between past actual monthly electricity demand against the historical three weather parameters (temperature, rainfall and humidity) and to present System Operators with a better model to do short term demand forecasting by inputting forecasted weather parameters which are available with forecasting agencies for few months ahead. Second objective of this study is to identify a better forecasting method in order to forecast monthly electricity consumption in Sri Lanka by comparing the forecasts done based on weather influence with other potential forecasting approaches.

The availability of valid, efficient and effective forecasting models for electricity loads is successful way of reducing the need for new generation capacity, scheduling resources, and by improving efficiency in electricity end use. In a broader economic sense, such models help in achieving a more efficient allocation of resources. Unlike long term expansion planning where the demand growth over 20 year time horizon is driven by external factors such as economic growth, population, tariff, etc, short term forecasting is primarily driven by weather and short term social factors such as holidays and festivals[1]. Due to higher living standards, weather has a considerable impact on the short-term electricity demand due to the use of fans, air conditioners and refrigerators [1]. There are three main weather related factors that affect demand. They are Rainfall, Humidity and Temperature [2]. At present, System operators (System Control Engineers) do the short term demand forecasting using time trend analysis and using their experience. Within a day, forecasting (which is done for shorter time steps of hours) is primarily done using historical hourly demand curves whereas weekly forecasting is done considering past data (in daily time steps). However, in system control center monthly forecasts (in monthly time steps) are done only based on time trend analysis. Therefore, in this study, it is anticipated to consider the weather variation in the country in forecasting the monthly electricity consumption.

Methodology

There are 62 CEB areas as covering the country and for these entire areas monthly tariff wise consumption and consumer accounts data were obtained for the period of January 2006 to June 2017 from Information Technology Branch, CEB. Monthly minimum and maximum air temperatures, minimum and maximum relative humidity and rainfall data along with latitude and longitude were collected from the Meteorology Department and Department of Census and Statistics for 20 meteorology stations as these stations are sufficient to cover the whole country. Under this study, starting from the simple forecasting methods up to complex methods various approaches were employed in forecasting monthly electricity consumption in Sri Lanka. They are,

1. Classical Approach of Time Series Analysis
2. Exponential Smoothing Method
3. Probabilistic or Stochastic Modeling Approach
4. Autoregressive Distributed Lag (ARDL) modeling approach
5. Artificial Neural Network

The original series of total monthly electricity consumption which was prepared by adding CEB area wise data, was divided into two parts as training and test set and the period of January 2005 to December 2015 was used to train the model for all forecasting methods while keeping the period of January 2016 to June 2017 (18 months) as the test set.

A time series is a collection of observations of some variable made sequentially in time, in which the time intervals are in regular intervals. Classical approach identifies four components (trend, cyclic fluctuations, seasonal variations and irregular variations) which affect time series values. Exponential Smoothing method is a very popular method in producing a smoothed time series. In single moving averages the past observations are weighted equally while Exponential Smoothing assigns exponentially decreasing weights as the observation get older. That means recent observations are given relatively more weight than the older observations when forecasting is done. The goal of the probability models for time series is to characterize random component of the series. The foundation of time series analysis with probability models is stationarity.

In order to deal with ARDL models and ANN models, weather variables were used as inputs. There were number of missing values in weather variables and to overcome this issue, missing values of weather parameters were interpolated through a computer program written in R (R version 3.4.1.). “imputeTS” is the only package on CRAN that is solely dedicated to univariate time series imputation and it includes multiple algorithms[3]. Among them most popular four algorithms were used and they are, Interpolation, Kalman, Seadec and Seasplit [4]. Performance of these algorithms can vary according to the structure of the series. Figure 1 contrasts the values imputed by these algorithms for a series of missing values of minimum temperature with actuals. To analyze the efficiency of the mentioned algorithms, its performance is compared with the root mean squared error (RMSE) which was chosen as a performance measure.

Figure 1: Actual values and imputed missing values by the algorithms

Model Building for Sub Regions in the Country

Sri Lanka as a tropical country has been traditionally divided into main climatic zones, namely Wet, Dry and Intermediate zones. Thus the weather variation is not uniform over the country. Hence unlike the first three forecasting approaches, the total monthly electricity consumption of the country cannot be taken as the response variable when influence of weather parameters on the consumption is in concern for ARDL and ANN approaches. Therefore, the need of dividing the country into sub regions such that each sub region possesses similar weather conditions was aroused.

As the simplest way, this categorization could not be done according to the climatic zones in the country because the electricity consumption was measured based on CEB areas. With this limitation the optimal way in order to get rid of this obstacle was matching the CEB areas into districts according to their geography boundaries. Then assigned the weather parameter values which were imputed for district centroids as each districts’ representative weather parameter values. Thus there are 20 districts and 3 CEB areas which were taken for region wise monthly electricity consumption forecasting.

In order to modeling the district wise and CEB area wise monthly electricity consumption, the meteorological parameters should be fair enough to represent the particular district’s or CEB area’s weather condition. Hence representative weather parameters have to be imputed for a specific location of each and every district and CEB area. Therefore as representative locations, centroids of the districts in Sri Lanka were found using Arc Map 10.3. The conversion of longitude and latitude into SLD99 systems was done for the 20 meteorology stations for the further applications in Inverse Distance Weighted (IDW) Interpolation and in order to extract the imputed values for centroids of districts. This was done using Arc Map 10.3. Due to sparseness of the stations, advanced methods such as “Kriging”, which accounts for the spatial correlation, is inappropriate. Hence distance based IDW interpolation was employed in estimating the weather parameters at district centroids. IDW is a spatially weighted average of the sample values within a search neighbourhood. In IDW interpolation method, the sample points are weighted according to distance and the influence of one point relative to the decreases with distance from unknown point which we are interested.

Figure 2: IDW Interpolation for average relative humidity for January 2006

Figure 2 represents the interpolation for average relative humidity for the month January 2006 and thus for all weather variables IDW interpolations were applied for all months (138 months). After identifying the relevant grid points, for each variable in each month the imputed values were gathered and those values are illustrated from Figure 3.

Figure 3: Time Series Plot of Interpolated Minimum Relative Humidity by districts

ARDLs are standard least squares regressions which include lags of both the dependent variable and explanatory variables as regressors. Then ARDL models were fitted to each region by incorporating the weather variables. In order to estimate the total electricity consumption in the country, ARDL models were built for 23 regions and by adding up all the forecasts given by each model, total forecasted consumption was obtained. In ARDL forecasts, there were two options as dynamic and static and compared to the dynamic forecasts, static forecasts which make the next step forecast using actuals was obviously better than dynamic forecasts where previous forecasted values taken in next step forecast.

ANNs are able to learn and model non-linear complex relationships. Unlike other prediction methods, ANNs do not impose any restrictions on the input variables[6]. Additionally, from some studies it has shown that ANNs are better even in modeling data with high volatility and non-constant variance. ANN was built to determine the performance of other forecasting approaches with its performance [7].

For the first three approaches total monthly electricity consumption in the country was considered as a univariate time series and forecasting was done while in ARDL model, since weather parameters were used, the district-wise and CEB area-wise consumption forecasting had to be done. Ultimately adding up all the forecasted values given by relevant ARDL models for each district and CEB areas, the total forecasted monthly consumption for Sri Lanka was calculated. Then the actual total monthly electricity consumption was compared with these first four methods. ANN approach was used for forecasting monthly electricity consumption in one region and it was compared with the relevant ARDL model for the same region in the purpose of comparison.

Results and Discussion

Under Classical approach and Exponential Smoothing method, additive models were the best model between those approaches’ Additive and Multiplicative models. Hence they were used for this comparison. Figure 4 graphically shows the forecasts by each four methods and the actual total monthly electricity consumption for the last 18 months which was undertaken as the forecasting period or test set.

Figure 4: Comparison of Forecasts of the Total Consumption with Considered Forecasting Models

Figure 5: Predicted vs. Actual of the Total Consumption for the Forecasting Models

According to the Figure 5, where the predicted vs. actuals by forecasting methods are presented in the same plot, it can be mentioned that the SARIMA model performs much inadequately compared to other three methods and the Additive model of Classical method and Additive model of Triple Exponential Smoothing models performs approximately in a similar way. Among the all four methods it was apparent that forecasts by ARDL approach is most probably performing better over other three methods.

Figure 6: Comparison of MAPE for the Forecasting Models

Accuracy measure MAPE for compared methods is graphically shown in Figure 6 and ARDL approach has the smallest value for MAPE. Hence it can be said that ARDL approach outperforms other three methods in forecasting. The performance of these forecasting methods can be listed in descending order according to MAPE as follows.

1. ARDL modeling approach
2. Exponential smoothing method
3. Classical approach
4. SARIMA modeling approach

As an overall, the weather variables used in this analysis are the imputed values by IDW interpolation, and therefore if it is required to forecast for a future month, then forecasts for weather variables provided by weather forecast agencies should be collected for these 20 meteorology stations and thereafter weather data should be imputed for centroids of districts by IDW interpolation. It can be presumed that if the actual weather data were available at district centroids ARDL and ANN models might be more preferable. The comparison between ARDL and ANN models was done using only one region. If it had been built 23 ANNs like in the case of ARDL modeling approach, via this generalization the performance of forecasting could be much better over all these forecasting methods. Thus it can be assumed that rather than a linear relationship there could be a non-linear relationship between the inputs and the response in this study. Therefore ANN can be suggested for further studies in forecasting monthly electricity consumption in Sri Lanka.

It was evident by the experience of system operators in SCC that the electricity consumption drops significantly in mercantile holidays. A dominant reason for this is that during the mercantile holidays industries are having a short break. However, in the ARDL models built for Trincomalee and Nuwara Eliya, number of mercantile holidays per month was significant but with smaller coefficients. A possible reason for this is that since both districts are known as tourist destinations due to migration during the holiday seasons, consumption of these areas could have increased.

This study was based on monthly electricity consumption and if it had been dealt with the daily consumption and weekly consumption, fluctuations of electricity consumption in mercantile holidays could be more sensitively captured. Although we have to deal with monthly data due to the limitations of obtaining data, weather influence on consumption also can be captured better if daily and weekly consumption were taken along with daily weather parameters.

It is a common fact that power failures are possible and due to that consumption may reduce, which has been neglected in this study. The LECO (Lanka Electricity Company Limited) sales are not included in the obtained data from IT branch of CEB and it was ignored under this study. The consumption data is updated to CEB data base through the meter reader. Meter reader visits and reports the consumption once a month and previous meter reading is subtracted from the current reading. That is how the monthly consumption is recorded as billing months and often meter readers visits happen in the middle of the month. Therefore the obtained monthly billing data might not be the corresponding accurate consumption for each month and it may include some portion of the actual consumption of previous month and the other portion from the actual consumption of current month. In future, CEB is going to establish smart meters with Subscriber Identification Module (SIM) and thereby the meter can be read remotely from CEB premises. With these data, forecasting can be developed further.

Conclusions

The main objective of this study was to find out the best forecasting model in order to forecast monthly electricity consumption in Sri Lanka and explore the influence of weather variables on the consumption behavior. Electricity consumption is highly correlated with the number of consumer accounts and weather variables. Correlation with weather variables depends on the monsoon season. Consumption of most of the districts had a moderate linear correlation with weather parameters. In the months of April and May, due to Sinhala and Hindu New Year festive seasons, the total electricity consumption was higher than that of in other months of the year.

Additive model performed better than multiplicative model in forecasting total monthly electricity consumption under Classical approach. In Exponential Smoothing method, the Holt-Winters’ Additive model was accepted as the best model compared to the Multiplicative model based on the measures of accuracy. model was the selected model from several candidate models in Stochastic approach. Since all the weather variables show a seasonal pattern, when missing values are distributed as a set of successive points, “Seadec” and “Seasplit” algorithms were always better in missing value imputation in weather variables. But in situations like one or two missing values have been distributed over the series, then any of the algorithms of missing value imputation was reasonable. Among all forecasting methods considered, ARDL model was the most appropriate model in forecasting the monthly electricity consumption. Also ANN models appear to perform even better than ARDL approach. Since both ARDL and ANN can be handled through other variables which are having influence on response variable, it can be concluded that rather than considering a single time series for the whole country it is better to deal as regression type models in forecasting monthly electricity consumption in Sri Lanka. Finally, it was found that weather variables have a considerable influence on the behavior of monthly electricity consumption in Sri Lanka.

References

1. Alex D Papalexopoulos and Timothy C Hesterberg. A regression-based approach to short-term system load forecasting. IEEE Transactions on Power Systems, 5(4):1535-1547, 1990.
2. James W Taylor and Roberto Buizza. Using weather ensemble predictions in electricity demand forecasting. International Journal of Forecasting, 19(1):57-70, 2003.
3. Steffen Moritz and Thomas Bartz-Beielstein. imputeTS: Time Series Missing Value Imputation in R. The R Journal, 9(1):207-218, 2017.
4. David S. Fung. Methods for the estimation of missing values in time series. Master’s thesis, 2006.
5. Ching-Lai Hor, Simon J Watson, and Shanti Majithia. Analyzing the impact of weather variables on monthly electricity demand. IEEE transactions on power systems, 20(4):2078- 2085, 2005.
6. Julian Faraway and Chris Chatfield. Time series forecasting with neural networks: a comparative study using the airline data. Journal of the Royal Statistical Society: Series C (Applied Statistics), 47(2):231-250, 1998.
7. TWS Chow and CT Leung. Neural network based short-term load forecasting using weather compensation. IEEE Transactions on Power Systems, 11(4):1736-1742, 1996.

Comparative Study on Soft Computing Approach in Weather Forecasting: Analytical Essay

Abstract– In a developing country, like India where the agriculture & industries are base for the national economy, the weather conditions play leading role for their proper development and smooth running. Therefore having accurate weather forecasting information may allow farmers or industry managers to make better decisions on managing their farms. Soft computing using ANN is an innovative approach to construct a computationally intelligent system that is able to process nonlinear weather conditions within a specific domain, and make prediction. A number of researches have been done or being done using Soft Computing Approach for forecasting. In this paper the presentation is all about to present the comparative study of several researches and some key findings that are initials for better start any soft computing model for prediction.

Keywords– Soft Computing, Artificial Neural Network (ANN), Back Propagation Algorithms, Multilayer Feed Forward Neural network (MLFFNN), Mean Square Error (MSE).

I.Introduction

Soft Computing is an efficient approach for forecasting, whether it is weather forecasting or any other else. A number of researches have been done focusing on the usefulness of soft computing approach in forecasting area. Here the presentation is utilizing some of researches to make some conclusions that are initials to be taken in account when planning forecasting using ANN technique. In traditional Weather forecasting approaches like:

• (a) The empirical approach and
• (b) The dynamical approach.

The first approach is based upon the occurrence of analogues and is often referred to by meteorologists as analogue forecasting. This approach is useful for predicting local-scale weather if recorded cases are plentiful. The second approach is based upon the equations and forward simulations of the atmosphere, and is often referred to as computer modeling. Because of the grid Coarseness, the dynamical approach is only useful for modeling large-scale weather phenomena and may not predict short-term weather efficiently. But for local scale & short term weather forecasting the approach of artificial neural networks (ANNs) is so efficient and a little bit easy.

ANNs provide a methodology for solving many types of non-linear problems that are difficult to solve by traditional techniques. Most meteorological processes often exhibit temporal and spatial variability, and are further plagued by issues of non-linearity of physical processes, conflicting spatial and temporal scale and uncertainty in parameter estimates. With ANNs, there exists the capability to extract the relationship between the inputs and outputs of a process, without the physics being explicitly provided. Thus, these properties of ANNs are well suited to the problem of weather forecasting under consideration. The popular soft computing techniques is ANN which performs nonlinear mapping between inputs and outputs, has lately provided alternative approaches to weather forecasting and so many researchers have taken in their research and come into the conclusion that ANN is best suited for forecasting. In the following paper the main objective is to find out some basic fundamentals and initials to make conventions about the ANN & forecasting.

Figure 1: Schematic diagram of data acquisition system

II. Implementation

To summarize the working of our system we can categorize the major components in following manner:

1. Data recording scheme
2. Parameter selection and user requirement definition
3. Neuro-fuzzy training and prediction scheme
4. Generating weather forecasting system

1. Data recording scheme:

The experimental setup consists of highly sensitive atmospheric pressure sensor and atmospheric temperature sensor . We have linked the sensors to the data logging computer using an interface cable to the parallel port LPT1. Data acquisition software records data in real time at a fixed time intervals of 4 seconds. Data precision for pressure sensor is 1 mbar and for temperature sensor is 0.1℃. System has capability to record the data continuously for several days without break in data recording.

2. Parameter selection and user’s requirement definition

We considered atmospheric pressure as a primary parameter and atmospheric temperature and relative humidity secondary type. Other parameters are also considered like wind direction and wind velocity.

Atmospheric pressure changes at any given place on earth are minute, hence, a very high sensitivity of pressure sensor is essential. Atmospheric pressure also changes with altitude and hence base atmospheric pressure differs from place to place depending on the altitude of the place from mean sea level. User requirement involves the collection of different input values and variables; including the selection of one or more forecasting measurements, the forecasting range, parameters such as the local forecast and distance of altitude from the sea level, and no. of samples per recording and interval of recording.

3. Neuro-fuzzy training and prediction scheme

Having collected and preprocessed all of the relevant weather information, our system starts the appropriate network training and forecasting, which is based on the back propagation in a neuro-fuzzy network. Table I shows the different categories defined for the fuzzification of the weather conditions.

The fuzzy data for predicting the occurrence of either rain or not rain ,bad weather ,good weather are based on the above given table I. While local pressure is measured in absolute scale , for simplicity of understanding of weather conditions pressure at sea level is computer with altitude information as correction factor to predict weather .

The following conditions are to be analyzed —

1. Constant low pressure (Bad weather)
2. Constant high pressure (Good weather)
3. Negative slope(Good to bad weather)
4. Positive slope (Bad to good weather)
5. Dual Slope (Variable condition of weather)

4. Generating Weather prediction

From recorded pressure data the generated weather forecast under stable low-pressure condition was rainy weather. As only two hours record was used for testing, we did not expect great results from it. Pressure changes recorded at fixed pressure rate of -0.564 mbar/h. It is very small change and is considered to be constant.

III. Results and Findings

After going through the above detailed study we find the following key results that play a major role in any forecasting model building. Apart from the traditional forecasting systems, ANN based forecasting is much feasible & best suited. Applying soft computing could be one of the best alternatives for local and short scale weather forecasting. The Study says any forecasting system using Artificial Neural Network & Back propagation Algorithms depends on following:

• The data: That we are going to acquire should be valid, Authentic and in proper format.
• The Variables: Means how many different kinds of data variables we use for input training set. As Temperature, Pressure, Relative Humidity, Due Point, wind speed, cloud status etc. The training results are dependent on the inter-relationship of these variables so it should be chosen so carefully and according to need. Many the variables, better the result.
• Data Analysis: These variables are interrelated & inter-dependent. So the interrelationships of these variables are a big factor in training set preparation & training of ANN. So the normalization should be done certainly & so carefully before making the training set.
• Dataset: The data that we acquire for training of our model plays a vital role in forecasting accuracy. This describes how much data we acquire for the training of proposed model.
• Training set: The training set is one the most considerable entity of our research work. Even it could be said the backbone of the ANN based forecasting system. It contains the input matrix & target matrix which contains the collection of unit input & unit output for the ANN correspondingly. The better the training set, better the result.
• Architecture of ANN: Once the training set is prepared the next most important thing to discuss is Architecture. The architecture of any artificial neural network is defined by the layers, numbers of neurons etc. Different forecasting models requires different forecasting architecture. The best suited ANN architecture for any forecasting model is the subject of research for a researcher. following are the key points to be taken in account when developing a forecasting model:
• Types of Network: The training and forecasting of any model is dependent upon the types of the network i.e. Multilayer Perception (MLP), Multilayer Feed Forward network (MLFFN) etc. in some cases MLP may suit best or in some cases it may be MLFFN. The appropriate type of network may converse fast for prediction.
• No. Of Hidden Layers: Our forecasting result‟s accuracy is highly dependent on the numbers of hidden layers. Some problem may converse in single layer ANN or some may converse in multiple layers ANN. Although single layer network is appropriate for solving any problem but it may be so less accurate. For better result multi layer network may be used that may converse slow but produce much better result. One drawback is that it goes complex. Multi layer Feed Forward neural Network (MLFFNN) is found the best for forecasting the weather condition.
• Algorithms: There are number of training algorithms are available but appropriate selection of training algorithms may leads fast and accurate forecasting results. Back propagation algorithm is found best suited with MLFFNN for forecasting the weather prediction.
• Activation function: A number of activation functions are available for training the network but selection the appropriate activation function may leads to better conversion. We can select any one of following TANSIG, LOGSIG & PURELIN.
• Weights/Bias: Apart from all above, another important factor is the initialization of weights and bias. Proper initialization of weights and bias may leads the network to converse fast and in proper direction.
• Learning Rate: Learning rate is another factor that leads the training of network with a given constant factor. High the learning rate fast the conversion & low the learning rate slow the conversion. A small learning rate may leads smooth conversion better results but slow and a high learning leads fast but less accurate result. Initially we may take any randomly value for learning rate.
• Threshold/Momentum: If we want an output by any particular condition then we can set threshold value .when the threshold value is achieved the output is generated else not. Another is the momentum that could be set for smooth conversion of network with the provided momentum factor.

Figure 2: Architecture of an ANN based model

In the Above shown figure 1 Inputs can be Temperature, Rain, Wind, Humidity, etc.

From conducted experiments we find the following key changes in the atmospheric pressure signature that can be related to dynamic states of atmospheric conditions and for meaningful short duration weather prediction. We are noticed the following key features in atmospheric pressure patterns that were related to the weather conditions and indicated trend of the future weather of the place. These conditions were:

• Stable day-night pressure gradient – indicator of stable weather -sun shine.
• Sudden pressure fall – indication of likely thunderstorm.
• Sudden pressure rise – indicator for windy day.
• Change in pressure slope – change of weather state in either way.

These trends were actually found in the recorded data and the selection of August month for the experiments was benefiting, as there were almost all types of signatures present in the atmospheric pressure indicating weather changes.

IV. Model Performance

Experiments were carried out on different locations of the different city. The performance of the experiments from stations having short difference from the sea level in the form of altitude is the best. These observations were carried out on the time series basis. The results for a single station or multiple stations do not produce a large difference in performance. The best result that is achieved by our model is due to the availability of a large amount of input data for the model to select the right variables. Thus the model has a greater chance of producing better prediction results .It is found and can be deduced that the correlation between the data at time t and t+1 is high; therefore, it is easier to build a successful model.

V. Conclusion

After going through all the above study & discussion we see that applying soft computing model for forecasting the weather conditions is most feasible rather than any other short term & local based weather forecasting approach. Through the implementation of this system, we illustrate how an intelligent system can be efficiently integrated with a neuro-Fuzzy prediction model to implement an online weather information retrieval, analysis, and prediction system by using electronic sensors. Increasing parameters for weather modeling may help in predicting weather changes to greater extend in comparison with than simple model used in present case.

References

1. https://www.sciencedaily.com/terms/weather_forecasting.htm
2. https://study.com/academy/lesson/methods-principles-of-weather-forecasting.html
3. Arvind Sharma and Prof. Manish Manoria, “A Weather Forecasting System using concept of Soft Computing: A new approach”, 2006; https://ieeexplore.ieee.org/document/4289915
4. Govind Kumar Rahul, Madhu Khurana, “A Comparative Study Review of Soft Computing Approach in Weather Forecasting”, 2012; http://www.ijsce.org/wp-content/uploads/papers/v2i5/E1053102512.pdf