Originally Posted by
h0gr1der
I'm going to throw out a comment that may be totally alien to some, some may have way more knowledge than me about it and be able to expound. We are all riding on the algorithm. The Spyder is computer controlled, and in some cases autonomously. I'd really like to see the actual findings in this case. It applies to all of us, in some way or another.
In the industry I worked in, I&E automation ran almost everything. Based on a hazard's identified danger to destroy or kill, and probability of an incident happening, IEC 61508/61511, SIL #1 through #4 dictated what type of control was mandated. A really bad thing that almost never happens gets a low SIL rating, and minimal control systems. A really bad thing that happens often gets a high SIL rating, sometimes mandating redundancy of control. Gibberish, I know, but let me continue.
Back to the Spyder. G force sensors, wheel speed sensors, yaw sensors, so many sensors and electronics. How many sensors have to agree before autonomous braking is applied? The wheel sensors appear to be Hall switches, either to an amplitude measurement input or high speed counter. To be safe, it should probably be a circuit with a built in design safety, commonly called closed loop. Some kind of reference signal (12 VDC), that if lost, causes the safe shutdown (Limp Mode) without causing an unsafe action (Full Braking Application). What if the one sensor that causes this kind of action fails in such a mode that it thinks it's OK, no loss of the feedback loop? If it's a single point measurement system (Low SIL Rating), then you get this kind of situation.
I'd really like to know what they find.