43a Ash Street Southport PR8 6JE[email protected]

Algorithms Used to Reduce Perishable Waste

If you are new to this area, we highly suggest reading out blogs on IOT, RFID, Algorithms, and Big Data Analytics.

Evidence highly suggests that waste may be a result of both insufficient data collection and handling. However, because RFID creates a more streamlined system for data collection and handling evidenced above, a faster and a more intricate response can be created where and when necessary, solving these issues. Moreover, this cannot be calculated or handled by humans alone, meaning algorithms are required.

When it comes to inventory management, traditional systems often consider shelf-life indefinite as discussed by Gürler et al (2008). Berk et al (2020) explain that more recent systems choose to use fixed lifetimes as this not only appropriate, but more successful in particular where a products lifetime is short and the ordering costs are low, it is suggested that it can reduce costs by more than half. Therefore, where lifetimes are short and ordering costs are high, the savings are substantial. Another important variable examined by Ueno et al (2007) are the environmental storage conditions where the products are stored. Producers and distributors have little control and often little knowledge of the environmental conditions their products are subjected to, to offset this a shorter expiration date than necessary is given. Jadermann et al (2009) discuss another benefit of RFID in that they can be used to track temperatures of individual products as seen in figure 1. Because of these factors, many perishables are wasted despite being usable.


Figure 1: Temperature within a sealed palette after 60-hour refrigeration (Jadermann, 2009).

Theory and research across many historical perishable inventory management systems clearly indicate that algorithms are needed to handle this issue. Indicated by Nahmias (1977), Graves (1982) and Ishii (1993) that agree the food supply lines of the world have consistently underperformed.  As the current data’s potential has not been optimized and a more sophisticated system would produce even larger volumes of data, this mean potential might not be reached without the use of waste prevention algorithms. A few of the algorithms available are discussed in this section, they are ABC analysis, least shelf-life first out (LSFO), and dynamic pricing examined by Wanitwattanakosol et al (2015), Qi et al (2014), and Chande (2005). ABC analysis is a simple but effective method used in inventory management where multiple products are stored. The algorithm splits the inventory into three groups, based on their annual cost volume usage.

  •         A product’s: high cost volume usage
  •         B product’s: intermediate cost volume usage
  •         C product’s: low cost volume usage

A summary of ABC analysis is provided by Hopp et al (1993) that state products in group ‘A’ should be ordered when they approach low volume to ensure that high costs products are not overstocked. Products in group ‘B’ require a continuous monitoring and replacement orders initiated when used. Whereas products in group ‘C’ require a two-bin system, meaning products should be divided into two piles, when a bin is empty it replenishes itself from the alternate bin and an additional order is placed to fill the previous bin. Where both bins empty before the replacement order arrives this indicates a stock requirement increase and vice versa. LFSO is an algorithm used within the core of computer systems and more importantly should be used where perishables occur at every stage of the worlds food supply lines. This algorithm uses a strategy which considers the expiration of the product and through this organises its availability in order to ensure that the products with the least shelf-life are used first as explained by Qi et al (2014). Chande et al (2005) discuss the many benefits of dynamic pricing algorithms, such as sales strategies being employed to move the unsatisfied demands. An example of this is where a product is overstocked or sales start to fall, advertising budgets can be targeted to move this product in real-time. All of these algorithms can be used in conjunction with one another and are greatly optimized when used in an environment where RFID provides real-time big data.

BERK, E., GÜRLER, Ü. and POORMOAIED, S., 2020. On the $$\varvec {} $$ policy for perishables with positive lead times and multiple outstanding orders. Annals of Operations Research. 284(1), pp.81-98.

CHANDE, A., DHEKANE, S., HEMACHANDRA, N. and RANGARAJ, N., 2005. Perishable inventory management and dynamic pricing using RFID technology. Sadhana30(2-3), pp.445-462.

GÜRLER, Ü. and ÖZKAYA, B.Y., 2008. Analysis of the (s, S) policy for perishables with a random shelf life. IIe Transactions40(8), pp.759-781.

HOPP, W.J. and SPEARMAN, M.L., 1993. Setting safety leadtimes for purchased components in assembly systems. IIE transactions. 25(2), pp.2-11.

JEDERMANN, R., RUIZ-GARCIA, L. and LANG, W., 2009. Spatial temperature profiling by semi-passive RFID loggers for perishable food transportation. Computers and Electronics in Agriculture65(2), pp.145-154.

QI, L., XU, M., FU, Z., MIRA, T. and ZHANG, X., 2014. C2SLDS: A WSN-based perishable food shelf-life prediction and LSFO strategy decision support system in cold chain logistics. Food Control38, pp.19-29.

UENO, K., HIROSE, T., ASAI, T. and AMEMIYA, Y., 2007. CMOS smart sensor for monitoring the quality of perishables. IEEE journal of solid-state circuits42(4), pp.798-803.

WANITWATTANAKOSOL, J., ATTAKOMAL, W. and SURIWAN, T., 2015. Redesigning the inventory management with barcode-based two-bin system. Procedia Manufacturing. 2, pp.113-117.

Leave a Reply

Your email address will not be published. Required fields are marked *