Nara Institute of Science and Technology

Title: Learning Food-arrangement Policies from Raw Images with Generative Adversarial Imitation Learning

Abstract: Food arrangement on a plate is one of the most challenging kitchen tasks for automation with robots. In particular, the food-arrangement planning problem has not been much studied to our knowledge, maybe due to the difficulty in its quantitative evaluation. In this talk, I introduce our attempt on the food-arrangement planning problem by an imitation learning approach from expert demonstrations. In particular, our approach employs a Generative Adversarial Imitation Learning framework, which allows an agent to learn near-optimal behaviors from a few expert demonstrations and self explorations without an explicit reward function. For evaluation, we developed a food-arrangement simulator for the Japanese cuisine “Tempura” with 3D-scanned tempura ingredients and conducted experiments for its performance evaluation. The experimental results demonstrate that our method can learn expert-like arrangement policies from bird-view raw images of plates without manually designing a reward function or requiring a massive number of expert demonstration data.

Ritsumeikan University

Title: Tactile Image Sensor for Food Inspection Tasks

Abstract: Detection of foreign substances mixed in food is one of the important issues in the manufacturing process of food factories. In this talk, we will discuss the use of tactile image sensors in food inspection tasks. A tactile image sensor consisting of a camera can acquire tactile information with high spatial resolution and can be used to detect small pieces of hard objects mixed in soft foods. In this study, we attempted to detect shrimp shells.

Ritsumeikan University

Title: Food Manipulation Competition Trials

Abstract: This video introduces two trials of food manipulation competition. The presenter organized food arrangement competition and wine pouring competition in my laboratory. In food arrangement competition, each team will drive a manipulator to pick food samples on the table and place them into a neighboring container. Each team designed a robotic hand attached to the manipulator and program the motion of the manipulator. The winner succeeded picking seven samples out of ten. In wine pouring competition, each team will drive a manipulator to pour the beads, which substitute for wine, into a glass fixed on the table. Each team will perform hand design and motion programming. The presenter obtained many issues to be tackled in the future competitions.

Iowa State University

Title: Robotic Cutting: Mechanics and Knife Control

Abstract: Home robots have long been a fascination to the public. They are at the core of the quality of life technology, carrying high promises for relieving people from daily chores, and providing costeffective health care to the growing elderly population and people with disabilities. Automation of kitchen skills is an important part of home robotics, and also one of the ultimate tests for robots to achieve human-like dexterity. Despite its significance and appeal, until today robotic kitchen assistance has been limited to dish washing and sorting, and to cooking of food items prepared by human.

In this work, a robotic arm equipped with a force/torque sensor is used to cut through an object in a sequence of three moves: pressing, touching, and slicing. For each move, a separate control strategy in the Cartesian space is designed to incorporate contact and/or force constraints while following some prescribed trajectory

Carnegie Mellon University

Title: Human-Robot-Food Interaction

Abstract: Cooking is a uniquely human activity. Thus, robot cooking competitions should include challenges that integrate humans. In this talk, I divide the process of cooking (and eating) into four phases, and introduce a challenge in each phase that involves human-robot interaction. By integrating human interaction into cooking competitions, I posit that we can drive forward research, expand real-world applicability, and increase public interest in our work.