Driving Simulator

We built our own simulator from original car seats and seatbelts
We are using a Logitech G27 steering wheel and pedals for control.
Currently, we’re using Genivi vehicle simulator using Unity to simulate driving scenarios.

During the driving session, we use the Biopac Student Lab software to record the psychophysiological signals of the driver such as EDA (Electrodermal Activity), ECG (Electrocardiography), or respiration rate. Electrodes are placed on the body and connected to a hardware module.
These signals can be analysed to assess the driver’s state during the driving session. 

♦ A new driving simulator has joined our lab for rapid tests or experiments that do not require the acquisition of psychophysiological data.
We have already tested the non-intrusive respiration sensor we developed

Scenarii

The software used for the simulation is the GENIVI open source software that is developed by a consortium of several companies, including some car manufacturers. It is developed under Unity and it uses some scripts in C#.

The software proposes 2 types of scenarios :

• Rural environment: the rural environment models Yosemite National Park. A 20 min autonomous driving is already available. It is also possible to trigger some obstacles such as falling trees, bears, rockslide, …
  We modified this scene in order to adapt it to the needs of our project. For this, we have implemented certain factors limiting the proper functioning of autonomous systems. These limitations are sloping road, adverse weather (heavy rain), fading lane marking, stationary obstacle (rock on the left lane) and a mobile obstacle crossing the road (deer) witch leads to a takeover request.
  The image below represents our test scenario in a rural environment:

• Urban environment: the urban environment models San Francisco city. The scene has many intersections and autonomous traffic is already implemented. It is also possible to trigger some obstacles such as pedestrians, dogs, stationary cars on the lane, …
  We are currently modifying this scene in order to adapt it to the needs of our project. The main improvements are to implement the autopilot and some limitations (unclear lane markings, external human factors such as pedestrians and bicycles, construction zones, …) in the scene.