Postdoc position : Upgrade and enhance the perception functionalities of Autonomous Ground Vehicles (AGVs)

Postdoc position at FactoLab:

Upgrade and enhance the perception functionalities of Autonomous Ground Vehicles (AGVs)

Factolab is a public-private laboratory between 3 public research laboratories at Clermont-Ferrrand (Institut Pascal, Lapsco and Limos) and the international company Michelin. The main goal of this common laboratory is to develop common research lines and technological solutions in the field of the “Factory of the Future” that can be applied in use-cases proposed by Michelin.

As part of the "Factory of the Future" initiative it is sought to upgrade and enhance the functionality of Autonomous Ground Vehicles (AGVs) employed at different sites of Michelin. The research work include replacing the existing sensors by solid state 3D LiDAR sensors and to develop algorithms/methods to implement different autonomous functions. These functions mainly include:

- Vehicle safety: detection of different obstacles in close proximity of the vehicle and classify them as potential threats to prompt a complete stop and ensure vehicle safety.

- Autonomous navigation: implementation of 2D or 3D SLAM methods to allow the AGV to map and navigate the factory floor with high level of precision.

- Handling of different objects: different objects commonly found on the factory floor need to be detected and transported to different designated places where they are carefully stocked with high level of precision.

The work is mainly divided in two parts; firstly real time simulation of the AGVs equipped with the designated sensors in different simulated scenarios. For this task, a dedicated simulator (from 4DVirtualiz - http://www.4d-virtualiz.com/produit_4dvsim/) would be made available to not only analyze the different sensor performances (configurations and characteristics) but also to emulate the different functions of the AGVs in different simulated scenarios (with increasing level of difficulty). The generated data from these simulations would be analyzed and used to implement and test the developed algorithms/methods for the different autonomous functions of the AGVs. After this validation phase, the second part consists of actually implementing the developed algorithms/functions on the real AGV.

The postdoc will work in collaboration with researchers from Institut Pascal (Dr Paul CHECCHIN, Pr Thierry CHATEAU and Pr Laurent Trassoudaine) and with some engineers from the Michelin research center in Lausanne, Switzerland.

Candidate should hold a Ph.D. in robotics (mobile robots) or a closely related field and should have a strong background and/or experience in 3D vision, 3D camera and lidar sensors, modeling and simulation, industrial real-time processes.

Duration: 12 months

To apply, send a CV, a cover letter, list of publications, best first-author papers in pdf format, and the contact details of 3 references to paul.checchin@uca.fr, thierry.chateau@uca.fr and laurent.trassoudaine@uca.fr