Racing car computers and wireless

Assuming one had a 100% reliable wireless connection, could one place the race-car’s computer(s) in the pit(to save car weight! How much?)

Would the turn around delay be significant?

I don’t know anything about the racing scene. I got this idea after reading indy 500 uses 802.11g for wireless communications.

I believe the biggest problem would be not enough bandwidth. According to this site, 802.11g is capable of a theoretical maximum 54Mbps, but about half of that is overhead. So, there is only 24Mbps usable. That is 3 megabytes per second under ideal circuimstances.

My gut feeling is that this is way too little bandwidth for real-time analysis of engine data. There’s simply a helluva lot of inputs and outputs in a racing car ECU. See the MoTeC M800 for example

The turn around latency would probably kill it. The computer is probably less than a pound, so it’s not like you’re saving that much weight anyway.

More info on wireless in racing.

I’m not seeing how you would really save anything. In order for wireless to work, you still need something to gather up all of the signals, convert them into a digital format, and transmit them out to the other end. Typically a tiny microcontroller is used for this sort of thing, but that’s pretty much exactly what the thinking part of an engine computer is, so you would just be swapping one microcontroller out for another one. Your net weight savings is going to be nil.

On top of that, it’s just not going to work because the response time of the system is going to be far too slow to be able to control the engine. Bandwidth isn’t so much an issue as the turnaround time is. The response time of a microcontroller is measured in microseconds. It’s going to take hundreds of milliseconds to transmit packets back and forth, which means your response time is too slow by several orders of magnitude.

You can do much better than “hundreds of milliseconds” with careful design. The PCM encoder can sample and digitize word N+1 while word N is being transmitted. On the receiving side, it isn’t necessary to wait for the whole frame (packet) to start processing it. Each word can be processed as it arrives. This reduces the total one-way delay to something on the order of three word-times, let’s say 240 microseconds on a 100 kbps telemetry link. That’s two serial/parallel conversions plus some additional time for delays in the transmitter and receiver.

Ok. I didn’t really think it was feasible, but just wondering.

IIRC last time I saw what someone told me was an ECU it looked like a small grey box about 2 inches x 2 inches or so and weighed about a pound.

How much do car computers (all of them together) weigh? How much do race-car computers weigh?(I assume they’re smaller)

Another oft-missed point in relation to professional motor racing is that most classes have weight minimums which racers are adding ballast to achieve.

That should probably be “many” not “most”.

Not really an apples to apples comparision. My car has 18-22 different computer nodes on the CAN system. With that said, race cars don’t have an audio system, nav system, SAT radio, climate, auto trans, ABS, Dynamic stability, Alarm, sunroof control, power windows, porwer door locks…get the picutre?
As has been said, an engine ECU is about a pound. FTR the driveline portion of the CAN system on my car runs at 500Kbs making the 100Kbs mentioned about 1/5 of the necessary speed for a street car.
What would be the competitive advantage of such a system? No weight savings, no advantage = no gonna happen.