I am exploring using Bayesian Networks to identify the best parameters within a system design, to improve its performance. I'm trying to find any case studies where this has been used successfully or unsuccessfully.
As an example consider an automatic assisted anti-collision system on a car. It could use some combination of ultrasonic, radar & IR seekers to sense and measure the closing speed with the car in front. These all have different costs / reliabilities and perform differently in different weather conditions. I am interested in understanding how effective Bayesian Networks can be in evaluating the trade space and identifying the best solutions to these types of problems.