A robust observer based algorithm to improve the accuracy of inventory measures in perishable supply chains

Managing perishable supply chains is a complex task that needs a precise knowledge of the current inventory level. Using a high-accuracy sensor equipment is the most intuitive and applied method. The related inconveniences are the large economic cost and the susceptibility to cyber attacks, like, e.g., in the case of radio-frequency identification technology. Moreover, even the most precise of physically realized sensors always provides measures affected by a certain amount of error. In this paper we show that it is possible to obtain a very accurate inventory information using a cheap sensor equipment. The proposed solution consists of the series connection of an estimator with a cheap physical sensor. The estimator exploits the noisy measurements provided by the sensor to monotonically increase the accuracy of the inventory estimate independently of the source and amount of the noise affecting the inventory measures.