Generating Fuzzy Coloured Petri Net Forecasting Model to Predict the Return of Products

Abstract

Currently various collection strategies are implemented in different parts of the world to effectively collect returned Managing product return is one of the issues in reverse products. The characteristics of those collection strategies logistics which has been developing with increasing enviare affected by the country's intrinsic factors such as inronmental awareness. Currently various collection come, education, types of environmental practices, enstrategies are conducted throughout the world to gather actment of WEEE regulation, and contribution of related old products to be reused, recycled or reprocessed In parties to the program [1]. Therefore, in order to establish order to effectively collect those products, these strategies an optimum collection strategy that mimics the characterhave to be designed to suit various locations. Therefore, istics of the city, state, province or country, different information on the quantity of product returns in those strategies have to be designed to cater for different localocations is essential. This paper presents an implementations. Consequently, it is important to know the rate of tion of Fuzzy Coloured Petri Net forecasting model to returned products over a specific time period in various predict mobile phone returns in various locations in Auslocations. tralia. The results shows up to 90 percent forecast accu-PRODUCT RETURN FORECAST racy in predicting mobile phone returns.A product can be returned or discarded because of several Keywords reasons such as the end of product life, technology develReverse logistics, product returns forecast, Fuzzy Colopment, sale of new product and many others. As disoured Petri Nets cussed by Rose et al [5] and Stuart et al [6] there are a number of factors which are affecting product life such as INTRODUCTION technology development rapidity, consumer demands for The emerging issues on environment and technology have improvements and product reliability. However, people triggered vast amount of research in reverse logistics. seldom discard their product immediately after it reached Reverse logistics was developed as a result of public sendof if naddtiopeopledont always recyeo awareness to conserve the environment. This awareness is send the p u additfon, people do not always recycle or translated into legislative action by government in the tomers' precyclin behvor . Moreover lau ten [7] enactment of Waste Electrical and Electronic Equipment sugestedytoincluesales.'of eroducto red old direcive b Euroean Uion [], th Wast Avoiance suggested to include sales of new p roduct to p redict old directive by European Union [1], the Waste Avoidance product returns. This notion follows that people tend to and Resource Recovery Act 2001 by New South Wales discard old products after they buy a new one. government in Australia [2] and electronic equipment andTherefore, the factors contributing to product returns can product stewardship act across different states in the US be summarized into three main categories: social, techno-[3]. Apart from the e...

Forecasting product returns for remanufacturing systems

“…Then, fuzzy controller was used to combine these impact factors and give one-period prognosis and neuro-fuzzy network was used to provide multi-period prognosis. Similarly, Hanafi et al [21] used fuzzy-colored petri nets to combine different sub-models, such as technology development, consumer demands, and product reliability to forecast returns at different locations over a specific time period. For others, non-parametric models are more suitable.…”

A Petri Net Approach for Green Supply Chain Network Modeling and Performance Analysis

“…The finishing product stored in distributor (P 3 ) is assigned to the retailer for selling (P 4 ). When the product is used out, it is collected (T 5 ) by the The raw material P 1 Material stored in supplier P 2 Manufacturer engages production P 3 The product stored in distributor P 4 Retailer sells the product P 5 Customer uses the product P 6 The recovery product inspected in collection center P 7 Repair/Refurbish of the product P 8 Cannibalization for remanufacturing of the product P 9 Material recycling of the product P 10 Disposal of the product P 11 Assigning recovery mode recycling collection center for detecting (P 5 ). Due to the fact that the customer using time of a product is much longer than other time consumption during the whole logistic flow, such as transporting time, manufacturing time, etc.…”

“…After generating the reachability tree, including tangible sates (the state only timed transitions are enable) and vanishing ones (the state at least one immediate transition is enable), the embedded Markov chain can be established, so does the calculation of stationary probabilities P[M i ], equal to the steady state distribution of tangible states. Transportation of the material from supplier to manufacturer T 2 Delivery of the product from manufacturer to distributor T 3 Distributor provides the product to retailer T 4 Customer buys the product from retailer T 5 The used product delivered to collection center T 6 The product is classified for direct reuse T 7 The product is classified for repair/refurbish T 8 The product is classified for cannibalization T 9 The product is classified for material recycling T 10 The product is classified for disposal T 11 The repaired/refurbished product send to distributor T 12 Delivery of the disassembled parts for remanufacturing T 13 Delivery of the recycling material T 14 Finishing disposal of the useless product T 15 The once reused product is classified for cannibalization T 16 The once reused product is classified for material recycling T 17 The once reused product is classified for disposal…”

“…Thierry et al [12] distinguished three classes of activities for the returned products: direct reuse/resale without reprocessing; product recovery activities including repair, refurbishing, cannibalization, remanufacturing and recycling; and disposal. Fleischmann et al [5] analyzed the general characteristics of recovery networks based on nine cases studies in different industries. Dowlatshahi [6] considered and analyzed the relevant literature in reverse logistics and generate a costbenefit analysis regarding reverse logistics.…”

“…Moore et al [11,12] investigated a Petri net approach in making disassembly process plans automatically in product recycling or remanufacturing. Hanafi et al [5] utilized a fuzzy colored Petri net in predicting the return of the end-of-life products in different locations for designing the collection strategies. However, these researches did not focus on integrity.…”