Digital Theses Archive


Tesi etd-03232021-155459

Type of thesis
A human/worker-centered framework and bio-inspired control strategies for foodservice robots
Scientific disciplinary sector
Istituto di Biorobotica - BIOROBOTICS
relatore Prof. DARIO, PAOLO
  • automazione per il servizio di ristorazione
  • chef
  • dataset
  • elettrodomestici professionali
  • flipping food
  • food service automation
  • food service collaborative robots
  • food tracking
  • force and torque
  • forza e coppia
  • grigliatura
  • grilling
  • intelligent systems
  • motion capture
  • movimento di girare cibo
  • professional appliances
  • robot collaborativi per la ristorazione
  • set di dati
  • sistemi intelligenti
  • tassonomia delle azioni
  • taxonomy of actions
  • tracciamento del cibo
Exam session start date
Restaurants, canteens, bars, catering – their main activity is to produce meals to customer order. Unfortunately, they face three critical issues: (i) unhealthy working conditions inflicted on their staff, (ii) shortage of qualified staff, and (iii) high turnover rates. The last two were previously associated to the first. Hence, foodservice workplaces need more autonomously-actuated equipment to support the workers’ physical tasks. This thesis argues that foodservice is a new environment for robotics research. So, its main goal is to initiate a new robotics field specialized in foodservice. Therefore, two tools are introduced: (i) the first taxonomy that makes the workers’ basic actions explicit, and (ii) a systematic review of the current mechatronic systems available for foodservice with implications for future decisions in robots development. Then, the thesis focus on the case study of grilling and the underlying action of flipping food, that is rarely studied. The goal is to understand the human techniques and, thus, collect kinematic and kinetic data. A custom experimental setup and protocol were successfully created to keep sensors safe and, simultaneously, record the execution of 2866 flipping movements with freshly cooked food, performed by 4 chefs and 5 home cooks. The approach can be reused to study other cooking actions and the collected dataset was released to public. Finally, a preliminary analysis revealed that some kinematic variables are adjusted to the food by both chefs and home cooks – a first insight to define bio-inspired control approaches for foodservice robots.