Artificial Intelligence for autonomous driving

The EIT Digital Doctoral School offers an Industrial Doctorate position to foster Autonomous Driving. The doctoral candidate will seek to apply Artificial Intelligence techniques to advance the deployment of Autonomous Driving with the support of EIT Digital Partners Centro Ricerche FIAT (CRF), Crysler Automobiles, and the University of Trento.

On board sensors in today's and tomorrow's cars provide a huge amount and variety of data. The challenge is to exploit these data. The Industrial Doctorate programme will lead the students to work on applying Deep Neural Networks, mimicking a human driver's brain, to make the most out of them.

Challenge

To accelerate the deployment of self-driving vehicles, the Automotive industry is hard at work at exploiting the latest advances in computer science to sort out the variety of real driving situations. Today's Advanced Driver-Assistance System (ADAS) supports the driver in avoiding driving mistakes, for example with a braking and lane departure warning system. With the help of on-board sensors, the car will have to "understand" its environment, as well as predicting what other traffic participants like vehicles, pedestrians, cyclists and animals might be doing. This is a real challenge, compounded by the need to internalise this understanding and take decisions in real time.

Approach

The approach pursued in this doctorate is to create a human-like vehicle in terms of its perception and reasoning capability, by leveraging on the latest Artificial Intelligence (AI) technologies.
Deep Neural Networks are being applied in a variety of areas to deliver a human like reasoning capability. Their application to the automotive field is still in its infancy. The Industrial Doctorate program aims at furthering their use, tackling issues specific of this domain, and addressing the:

  • Fusion of heterogeneous data (for example self-calibrated estimation of vehicle state with transparent fusion odometer, gyro meter, GPS video and Lidar).
  • Use of deep neural networks to mimic loops in the brain (such as action selection to deal with multiple choices that exist when there are options).
  • Prediction of expected perception by using network backward approaches (for example depict the road behind an obstacle to imagine unexpected situations).

Expected outcome

The PhD research is expected to apply AI technologies to deliver one or more innovative applications in the automotive domain. Expected research results also include algorithms and numerical tools for AI system training, adaptation and performance evaluation.
These expected outcomes are aligned with the efforts of the Digital Cities Innovation activities in which EIT Digital keeps investing, as pointed out by Stéphane Péan, Digital Cities Action Line Leader: "We focus on urban mobility and, particularly, on autonomous vehicle solutions. The proposed PhD topic is very well aligned with this priority. In the next decade, Artificial Intelligence will have a tremendous impact on the design of connected and autonomous vehicles, which are supposed to revolutionise urban transportation systems in our cities."

Location

The position is based in the Doctoral School Training centre in Trento where a strong ecosystem for Digital Cities exists. The PhD student might have the opportunity to spend part of the mobility programme in places like the University of Sheffield to look at research loops of neutral networks organised as in the human brain and the Middlesex University in London where studies on subsumption perception-action architectures are well advanced.

Facts

  • Industrial partner: Centro Ricerche FIAT (CRF)
  • Academic/research partner: Università di Trento
  • DTC location: Trento
  • Number of available PhD positions: 1
  • Duration: 3 years
  • This PhD will be funded by EIT Digital, The University of Trento and CRF.

Apply

Those interested in applying should fill in the application form on the website of The University of Trento.

Please apply before August 31, 2017.

Read the full position paper here

© 2010-2019 EIT Digital IVZW. All rights reserved. Legal notice

Supported by the EIT