Bachelor and master students in Mechatronics are involved in project “DEEPCOBOT”. The projects are supervised by Prof. Jing Zhou and Assoc. Prof. Ilya Tyapin. The bachelor and master students have worked together with PhD students Emil and Jayant.
This article is more than two years old, and may contain outdated information.
The objective of this bachelor project is to provide a solution to perform semi autonomous collaborative tasks, between a robot and a human. The project is focused on making a usable autonomous solution, that will find a sequence of actions that will lead to a desired goal. The tasks include solving different planning problems such as identifying objects, picking and placing objects, moving around in a known environment, and interacting with the environment to complete its goal. To achieve these tasks it was necessary to make new software for the robot to run. All programming was done within the ROS framework. By using ROS melodic, Linux Ubuntu 18.04 and Python programming language. The majority of all testing was in simulation using Gazebo, testing was done on the physical robot when simulation was satisfactory.
Deepcobot is called “Collective Efficient Deep Learning and Networked Control for Multiple Collaborative Robot Systems”, which is a multidisciplinary frontier research project funded by the Research Council of Norway (in total 20 million NOK). The project period for deepcobots is from 2020 to 2025.
Professor Jing Zhou is the project manager. She is leading and coordinating the project, a collaboration between UiA and the partners, Mechatronics Innovation Lab (MIL), ABB Norway, Omron Electronics Norway, the University of California San Diego (USA), KTH Royal Institute of Technology (SWE) and the University of Navarra (SPA).
Emil Mühlbradt Sveen and Jayant Singh are PhD students in Deepcobot. In total, there are three PhD students and one postdoc.
In recent years, robot systems have received a great deal of attention in the context of a large number of industrial applications, such as additive manufacturing, automotive manufacturing, material handling, packaging and co-packing and quality inspection. The collaboration between multiple robots and human operators is considered to be the most prominent strategy in Industry 4.0 and future Industry 5.0, sharing the same space and collaborating on tasks according to their complementary capabilities. There is rising demand for robots to solve complex tasks in industrial companies in Norway.
Deepcobot project will investigate the design of a new generation of decentralized data-driven Deep Learning based controllers for multiple coexisting collaborative robots (Cobot), which interact both between themselves and with human operators in order to collectively learn from each other's experiences and perform cooperatively different complex tasks in large-scale industrial environments. This is motivated by the increasing demand of automation in industry, especially the demand of a safer and more efficient collaboration between multiple Cobots and human operators to integrate the best of human abilities and robotic automation.
The vision of this project is that the learning of the optimal local control policies can be substantially accelerated by sharing both information about previous experiences and computation across multiple neighbor Cobots connected through a wireless communication network, providing solutions that satisfy the necessary real-time constraints in the considered robotic applications, as well as providing sufficient robustness and interchangeability to the control solutions. This multidisciplinary project covers the areas of deep learning, optimization, reinforcement learning, decentralized shared control, embodied Artificial Intelligence (intelligent robots and devices), bi-directional interaction between Cobots and human operators, and cross-layer networking with a significant potential in industrial applications.