The proposed subject is made in the context of the IMARA team framework about mobile robotics, wireless communications and driving automation.

Today, vehicles become more and more autonomous. Many Advanced Driver Assistance Systems (ADAS) are embedded in order to help the driver in the driving process. This is possible since vehicles are equipped with many sensors. Proprioceptive sensors (acceloremeter, gyrometer,…) provide information about the vehicle by itself such as its velocity or lateral acceleration. On the other hand, exteroceptive sensors, such as video camera, laser or GPS devices, provide information about the environment surrounding the vehicle or its localization. As data are noisy, inaccurate and can also be unreliable or unsynchronized, the use of data fusion techniques [Hall, 1997] is required in order to provide the most accurate situation assessment as possible. In other words, situation assessment consists in providing a local map, modeling the vehicle state by itself (position, velocity, acceleration, braking ability,…), but also potential obstacle states (position, velocity, type,…) like other vehicles, bicycles or pedestrians, and finally the environment state including weather conditions or road state. Sensor data can be also merged with information coming from knowledge database, such as Geographical Information System (GIS) containing terrain elevation map or road coordinates.

In automated or highly assisted driving modes, localization of our own vehicle in an absolute referential is crucial. It has been studied since many years [Blackman, 1999]. Techniques are more and more sophisticated, taking into account more elaborated motion models in Multiple Model approaches [Blom, 1988], integrating proprioceptive data [Gruyer, 2011] as well as road coordinates [Kirubarajan 2000] or using information coming from communication [Pollard, 2012]. Study of the literature shows that the tendency for ego-localization is to use multi-sensor approaches, but it is often limited to the use of two heterogeneous sensors. As an expert in data processing and fusion, IMARA already developeded several localization approaches. The goal of this PhD is thus to use building blocks in order to create a multi heterogeneous sensor system in order to improve accuracy and reliability whatever the conditions. The PhD should focus on the way to deal with uncertainty and breakdowns by creating a multi-modal system. He should start by creating the most reliable localization algorithm as possible, by using all available data (GPS, map, IMU,…), but with real time constraints. Then, he should deal with deteriorate data and create several reaction modes. Many scenarios can be imagined: degradation of the map accuracy or erroneous information, lost of the GPS signal, breakdown of one sensor, … in order to keep the best accuracy as possible.

The scientific issues would thus mainly focus on estimation theory and data fusion. In addition, in order to validate the theoretic work developed during this PhD, a real-time demonstration could be proposed to demonstrate the system realism

Requirements:

Requirements include a completed M.Sc. degree in electrical and computer engineering or equivalent, solid background in mathematics and statistics, as well as a good background in scientific software development and programming (C++, Matlab). Knowledge of English is mandatory (French is a plus) and teamwork skills are required.

This position if offered and located at INRIA-IMARA

Centre de recherche Inria Paris – Rocquencourt
Domaine de Voluceau – Rocquencourt

B.P. 105 – 78153 Le Chesnay Cedex

For inquiries and applications, please send a resume and a cover letter to:

Dr. Fawzi NASHASHIBI (fawzi.nashashibi@inria.fr) , Head of the Imara team

0033.1.39.63.52.56

or

Dr. Evangeline POLLARD (Evangeline.pollard@inria.fr)

0033.1.39.63.53.26

References:

[Hall,1997] D. Hall and J. Llinas, “An introduction to multisensor data fusion,” Proceedings of the IEEE, vol. 85, no. 1, pp. 6–23, Jan 1997.

[Blackman, 1999] S. Blackman and R. Popoli, Design and Analysis of Modern Tracking Systems. Artech House, 1999.

[Blom, 1988] H. Blom, Y. Bar-Shalom, “The interacting multiple model algorithm for systems with Markovian switching coefficients ”, IEEE Transactions on Automatic Control, vol. 33, no 8, pp 780-783, 1988

[Gruyer, 2011] D. Gruyer and E. Pollard, “Credibilistic imm likelihood updating applied to outdoor vehicle robust ego-localization,” in Proceedings of the 14th International Conference on Information Fusion (FUSION), July 2011, pp. 1 –8.

[Kirubarajan 2000] T. Kirubarajan, Y. Bar-Shalom, K.R. Pattipati, I. Kadar: “Ground target tracking with variable structure IMM estimator”, IEEE Trans. On Aerospace and Electronic Systems, Vol. 36, No. 1, pp. 26-46, Jan. 2000.

[Pollard, 2012] E. Pollard, D. Gingras, “Improved Low Cost GPS Localization By Using Communicative Vehicles“, ICARCV, Guangzhou (China), 2012