2021 IEEE 27 th International Symposium for Design and Technology in Electronic Packaging ( SIITME) 27-30 Oct 2021, Timișoara, Romania Digital Solutions for Smart Food Supply Chain George Suciu R&D Department BEIA Consult International Bucharest, Romania george@beia.ro Serban Calescu R&D Department BEIA Consult International Bucharest, Romania serban.calescu@beia.ro Iulia Pop R&D Department BEIA Consult International Bucharest, Romania iulia.pop@beia.ro Robert Vatasoiu R&D Department BEIA Consult International Bucharest, Romania robert.vatasoiu@beia.ro Adrian Pasat R&D Department BEIA Consult International Bucharest, Romania adrian.pasat@beia.ro Ioana Suciu R&D Department BEIA Consult International Bucharest, Romania ioana.suciu@beia.ro Abstract—Food waste is a major problem in the current economy, equaling 20% of all food produced in the EU. The European Commission reported, in 2019, that the EU-27 generated 89 million tons of food waste, while the food service and hospitality sector accounts for 12.5 million tons of food waste (14% of total food waste). Catering and Food Services in Healthcare institutions face even more significant challenges: plate waste in healthcare facilities ranges between 6% - 65%; hospitalized patients continue to suffer from the risk of malnutrition resulting from inappropriate food distribution and consumption. In this paper we introduce ADCATER project, which proposes an advanced ICT platform based on innovative technologies integrated in a Smart Food Catering Supply Chain platform. ADCATER aims for economic efficiencies to farmers and suppliers; personalized nutrition accuracy to diners/kitchens, wholesalers, and caterers; also, verification of personal, organizational, and global nutritional policies. This will be achieved by harnessing IoT, computer vision and deep learning technologies to identify and decode images of prepared food “served to plate before a meal” and “left in plate after a meal”, applying advanced analytics to derive valuable information such as served meal ingredients, the degree of adjustment and dietary gaps to the diner profile, batch traceability data, effective and up-to-date nutritional supervision, food waste, and correlation between consumer consumption and health. Keywords—food waste, blockchain, AI, computer vision, supply chain I. INTRODUCTION Preventing and reducing food waste has become a global priority. One-third of the food produced in the world is wasted every year [1], 20% of the total food produced in the EU is lost or wasted, and 43 million people cannot afford a quality meal every second day [2]. Food Processing and Food Services generate almost a third of food waste in Europe (2nd after Households, which generate over 50%) [3]. Catering and Food Services in Healthcare institutions face even more significant challenges: plate waste in healthcare facilities ranges between 6% - 65%; hospitalized patients continue to suffer from the risk of malnutrition resulting from inappropriate food distribution and consumption [4]. The prevalence of malnutrition in the hospitals is close to 35%- 40%, causing increased healthcare costs, prolonged length of stay, and unfavourable prognosis [5]. Food traceability systems (FTSs) reveal inefficiency in either material or information flow within an enterprise or between supply chain partners. Thus, one of the key parts of FTSs regards the implementation of IoT solutions to gather data by monitoring multiple environmental parameters to analyse the crucial stages of the food production phase over the crop cultivation systems [6]. In this paper we present the ADCATER’s main proposed advances beyond existing food supply chain traceability approaches, in Section II we describe the methodology for developing the smart food catering supply chain platform, while in Section III we present the preliminary results regarding (i) the first layer of the supply chain, IoT data acquisition, and (ii) the data traceability module based on blockchain technology. Finally, in Section IV we summarize the ADCATER framework concept and draw the conclusions. II. METHODOLOGY The following subsection will present the core concept of the ADCATER’s framework and architecture elements. In subsection two we describe the case studies through which our platform concept is validated. A. DDSN Methodology ADCATER implements a Demand-Driven Supply Network (“DDSN”) methodology for developing a smart “farm to fork” platform that measures and analyses data from all segments of the preparation and consumption chain, providing accurate information to ensure proper nutritional consumption by diner/patient, and economic efficiencies for management. The ADCATER’s main proposed advances beyond existing food supply chain traceability systems are presented in Fig. 1: (1) The integration of AI and control elements linking the food consumption with candidate origin along the production and supply chain directly; (2) AI methods to estimate the consumed food and not only served food; (3) Real-time monitoring of the entire supply chain to detect future shortages. The main end-users are catering services in hospitals, senior-care centers, schools, or businesses, which