Development of an Automatic Control Platform for Food Waste Management

• Main Author: Modin, Ivan
• Format: Others
• Language: Swedish
• Published: Uppsala universitet, Institutionen för elektroteknik 2021
• Subjects: automation; hållbarhet; produktutveckling; Robotics; Robotteknik och automation
• Online Access: http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-451933

According to the facts presented by The United Nations in their 12th goal - ensure sustainable consumption and production patterns - of the 17 sustainability goals, an estimated third of all food produced each year ends up wasted, in the bins of consumers and retailers or spoiling, due to poor transportation and harvesting practices. Reducing food waste is therefore one of the indispensable tasks to accomplish this goal.   This degree project aims to develop, as proof of concept, an automatic control platform for food waste management and implement it in a form of a prototype. The platform is intended to handle the workflow automatically, taking in food waste, weighing it, getting it photographed for image recognition and transporting it to a specified bin according to the classification of food waste using image recognition. The classification part is, however, not included in this project.    The prototype of the control platform consists of a mechanical construction, a weighing scale, a stepper motor, a servomotor and a Raspberry Pi single-board computer. The mechanical construction was designed using CAD (computer aided design) and made with a 3D printer. On the top of the construction is an open flat platform, for placing food waste, which has two walls on the sides, one opening in the front and a robot arm in the back. The robot arm is moved forward, to offload food waste, and backward by a servomotor. The platform is mounted a base with a stand which supports the mechanical construction. It is rotated using a stepper motor installed in the base and stops at three different postions at 90, 180 and 270 degrees, respectively, which are to represent the three different categories of food waste that is to be classified using image recognition. The Raspberry Pi is the controller that is programmed to coordinate the whole workflow.   The prototype is finished and tested. The test results have shown that it works well as a control platform that can handle the workflow, starting at the position to take in food waste, then weighing it and transporting it to three different specified positions, respectively.