Programming an industrial robot is complicated and costly. Currently around CHF 80’000 is needed just for programming and hardware integration. This results in more than 75% of industrial tasks being not automated. Additionally, due to the absence of standardization in the way robots are programmed, current solutions are inflexible and difficult to implement. Thus, programming requires strong expertise in robotics, which most employees do not have, despite working closely with or around robotic systems.
The LASA EPFL laboratory has pushed the boundaries of research in collaborative robotics and developed a framework that simplifies the control and programming of robotic tasks. The solution combines machine learning and advanced robotic controls to ensure the safe execution of the planned action, even in changing and uncertain environments. Not only is this crucial for the development of collaborative robot applications, but it also benefits industrial robotics by adding flexibility in the traditionally rigid robotic application implementation. AICA, a spin-off of the LASA laboratory, aims to package the novel software and deploy it in such a way that any worker onsite can program, supervise and modify the execution of a robotic task.
AICA is tasked with packaging and deploying the robotic control framework that is the outcome of the research undertaken at LASA. It has been awarded with an Innogrant support fund and successfully attained VentureKick stage 2.
As part of the Innobooster project, the focus will be made on developing a portable demonstrator of the software in the form of a graphical interface allowing the programming of a simulated robot. This demonstrator will allow the company to prospect potential clients during factory visits, and sign first proof of concept projects. A partnership with a key industrial player is underway (name under NDA) and the Innobooster fund will help us tackle the first developments with this partner. A second objective of the project will be to release the control software in an alpha version for our first clients.
Am Projekt beteiligte Personen
Letzte Aktualisierung dieser Projektdarstellung 11.10.2021