Cognitive Modelling Tools for Simulation Studies in Semi-Autonomous Vehicles

Pallav Rawat
Aug 24, 2017 · 3 min read

Modelling a Semi-Autonomous or an Autonomous vehicle in a virtual environment has proved to be very successful lately for various reasons. Firstly, it facilitates data collection at much lesser risk as compared to on-road testing. Secondly, it is less costly and acts as a complete proof-of-concept before you buy all the expensive Inputs/Outputs for your system and get the real prototype on the road. At the same time, present day virtual environments are not completely fool-proof and surely cannot mimic advanced road situations.

This switching from autonomous driving to manual mode and vice versa will be an essential part in Level 3 Autonomous vehicles. Studies at Stanford have proved that taking back control from self-driving mode affects human steering behavior. There are other issues as well, like drowsiness, loss of attention etc. which leads to compromised performance. With semi-autonomous vehicles being pushed onto the market, driver monitoring in virtual simulation systems play a relevant role in determining the user’s fitness to drive whenever an automated-to-manual transition of control occurs.

Possible assessment tools for cognitive modelling of human behaviors in risky road situations are:

(1) Electroencephalographic (EEG) Activity Monitoring: An EEG measures electrical activity of the brain which can be plotted with driver inputs for concentration assessment. EEG in depth is explained here. A useful plot would look like :

Mental workload is synchronized with driver controls. Courtesy: REVS program at Stanford

(2) Skin Conductance(SC)/ElectroDermal Activity(EDA) : These sensors can be used for several biofeedback applications and typically measures sweat gland activity which is closely related to arousal and stress. Among the various skin conductance analysis options, Ledalab is one and can be modelled on MATLAB.

(3) Eye Tracking : Driver fatigue has proved to be one of the major reasons of road accidents. Eye detection in a simulation model could help to model driving patterns useful in detecting drowsiness and this in turn, would alert the driver. This can also be used against use of mobile phones while driving, one of the major driver distraction issues faced today. This tool is already used by major car-makers as an Advance Driver Assistance System(ADAS). Top eye tracking companies are ranked based on their number of publications and listed here.

Time to take a break!

Some Useful Links :

)

Pallav Rawat

Written by

Passionate about Autonomous Cars and Data Science

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade