Postdoc position Robot Manipulation in Retail Environments

Robots are slowly entering the retail domain (supermarkets, stores, warehouses, delivery, etc.). In this project we focus on mobile manipulation in retail environments. Your work will focus on grasping and manipulation.

The PostDoc position is in the context of the project "AI for Retail Lab" (AIR-Lab) Delft https://aiforretail.ai/ , a joint TUDelft-Ahold Delhaize industry lab. Ahold Delhaize is a Dutch multinational retail company active in the Netherlands, Belgium, the Czech Republic, Greece, Luxembourg, Romania, Serbia, and the United States. Besides supermarkets, Ahold Delhaize is also active in the fields of liquore stores, drug stores, and as online grocer, as well as the largest webshop in the Netherlands.

You will be working closely together with a team of 4 PhD students and 1 PostDoc (covering topics such as whole body motion planning, self-adaptive robot architectures, design of human-robot interactions, optimization for last mile delivery, and ethics) as well as a team of technical support staff. The PIs in AIR-Lab Delft are Javier Alonso-Mora, Carlos Hernandez-Corbato, Jens Kober, Marco Rozendaal, Filippo Santoni de Sio, and Martijn Wisse with all of whom you will have regular interactions.

The handling of bulk units can be largely automated, handling individual items is still an open problem despite tremendous progress in pick and place actions. For grasping and manipulation, particular challenges include:

• handling fragile objects, e.g., fruits and vegetables that not only exhibit a large natural variation but are also flexible and fragile.
• handling a huge variety of objects, e.g., supermarkets carry a few tens of thousands items.
• handling narrow and tight spaces, e.g., items are tightly packed on shelves, requiring pulling / pushing them for order picking / re-stocking, but also perception is limited.

Rather than manually designing and programming robots to deal with these challenges, we believe that robots will need to be equipped with the capability of adapting and learning on their own, i.e., with artificial intelligence. Besides mainly data-driven methods such as (deep) reinforcement learning, inverse reinforcement learning, or imitation learning, we are also very interested in methods that integrate prior knowledge (e.g., in the form of models, planning, reasoning) to render learning processes tractable and efficient.

The main focus of the Cognitive Robotics department is the development of intelligent robots and vehicles that will advance mobility, productivity and quality of life. Our mission is to bring robotic solutions to human-inhabited environments, focusing on research in the areas of machine perception, motion planning and control, machine learning, automatic control and physical interaction of intelligent machines with humans. We combine fundamental research with work on physical demonstrators in areas such as self-driving vehicles, collaborative industrial robots, mobile manipulators and haptic interfaces. Strong collaborations exist with cross-faculty institutes TU Delft Robotics Institute and TU Delft Transport Institute), our national robotic ecosystem (RoboHouse, Holland Robotics) and international industry and academia. http://www.cor.tudelft.nl/