Evaluating Driver's Behavior

Latest Advanced Technology and Tasks in Automobile Engineering

Seminar 2: Observation and Evaluation of Driver’s Behavior

Jenna M. Eason

 

1. Look and touch the sensor car.  Observe what kind of sensors attached.  What sensors do you think yet need and why?

I believe there is a possibility of a device that can observe facial expressions and quantify this data.  For example, a static picture of the driver’s face can be taken in a relaxed environment.  Video will then record deviations in expression, notably in the eyes, 

brow, mouth, and possibly clenching of the teeth.  As a driver gets excited, upset, or drowsy, these expressions will be detected and analyzed.  For example, the level and angle the brow raises or droops is proportional to the level of the driver’s excitement.  Of course this data will vary greatly with different drivers, but the delta is what will be plotted.

Another sensor could record brain activity and map this data.  This information shows excitement vs. lack of activity and is a more direct measurement the physical actions.

A final sensor could detect the amount of pressure of the driver’s back against the seat.  If the driver is excited they may lean forward to see the road better and grip the steering wheel more tightly (another possible sensor) or press against the back if they are slamming the breaks.  This data is to be visualized next to normal.

A final thought is another option for test conditions.  The driver should be tested when eating and smoking to detect how these actions affect driving ability.

2. Plot a car trajectory of taking over.  Describe which part of the trajectory representing the driver’s characteristics and how.

A. Driver is calm, releasing pressure on the gas as he prepares to predict passing the vehicle in front.

B. Driver leans forward and begins to excite when he sees car in passing lane that was not visible from A.  Driver must accelerate.

C. Driver is very excited, palms are sweating, talking has completely ceased as he tries to maneuver between cars.  At the end of C driver has to break unexpectedly because of front car and is very excited.

Driver decides not to pass anymore after this excitement.  He is a very bad driver for trying to make this pass.

3. What do you think is a good measure to detect drowsiness?

Facial expressions are key to detecting drowsiness.  The same device described in Question 1 will work well.  Signals for drowsiness include droopy or blinking eyes, yawning, fidgeting, head dropping, and body position.  The control over the vehicle will decrease and the vehicle will start to swerve.  This motion can be recorded and if repetitive can signal an alert.   Also, often when drowsy, the driver will unintentionally “tailgate” the vehicle in front.  This can be recorded with the radar detection already placed on the front of the vehicle.  If there is mapping for the brain, a wandering mind is a sign of drowsiness

4. Describe overall impression of the class.

I thoroughly enjoyed this seminar and am very interested in detecting driver behavior and emotional response to improve the driving experience.