AIM Solo predictive lap time algorithm

[Evergreen Dan]

Does anyone understand/know exactly how the AIM calculates it's predictive lap time? I couldn't find it documented anywhere.

- I believe it uses the best lap of the current session (since power-on) as the reference lap.3
- I don't believe it just calculates where your current lap is in comparison to the reference lap and assume that you will complete the rest of the lap at a comparable relative speed (i.e. x% faster). I say this because sometimes the display changes by a few seconds at S/F and if they were doing this, the actual lap time and predicted lap time would be identical just before passing S/F since the whole lap is history.
- I wonder if it breaks the lap up into segments based upon g. I sometimes see the display change right after a turn.

Anyone know?

[aeronaut]

At first I thought you were asking about 'best possible lap', but that's easy.
So, no, I have no idea, but good question.
I'd email the guys at AiM, they've been extremely responsive and helpful when I've emailed them.
The last Race Analysis release even had one my recommendations as an update!!!!

[CoMZ302]

Does anyone understand/know exactly how the AIM calculates it's predictive lap time? I couldn't find it documented anywhere.

I'd like to know this as well.

I do know, in RS2, they have two calculations you can use for best lap: one is 'best rolling lap' and the other is 'best theoretical lap.'

Best rolling lap is really great - finds the fastest connected series of segments (across lap boundaries), and simply moves start/finish to whatever achieves that. It's far less theoretical as you did actually in fact run that time (if they moved start/finish to wherever was appropriate).

The other calculation is far less realistic. If you go out for 10 laps, it finds the best time through each segment, regardless of which lap you were on, then adds them all up. While you did in fact run each segment time, you may not have been able to do so without breaking up the lap - for example on a hot lap you may be faster by slowing down through a segment to set yourself up for the next; this calculation ignores that and includes the one time you flew through that segment which screwed up the next one. In short, it may not be physically possible to achieve the number this one comes up with...

...that said, you should be pretty close to that lap with your best lap. If you're 1+ seconds off, you got some drive'n to do. (as I do!)

Anyway, it's probably using segments, and adding up what you've done so far, with the segments from.... yeah, that's the question. What is it guessing will be your segment times for the parts of the track you haven't gotten to? Your last time through those segments? Best times? Best lap segment times?

I DEMAND ANSWERS

[Evergreen Dan]

I wrote to AIM. Will report back.

[CoMZ302]

Anything?

[Evergreen Dan]

Oh, sorry. Yes, I got an answer. No, it didn't say anything informative.

I can't tell you exactly how it is computed. The real-time ± is based off best lap in the session and track position. The predictive lap time, showing the actual pace of the lap you're on, is updated every few hundred feet, again based on track position. I hope this helps.

I'm still not sure why the predictive lap time is sometimes (not often) off by a couple of seconds just before crossing S/F. I guess I'll just have to not overthink it.

[aeronaut]

Oh, sorry. Yes, I got an answer. No, it didn't say anything informative.


I'm still not sure why the predictive lap time is sometimes (not often) off by a couple of seconds just before crossing S/F. I guess I'll just have to not overthink it.

How often does this happen? As in, when this happens could it be a GPS reception problem (aka not knowing accurate track location)?

[NotoriousJJE]

Thi has never happened with mine btw.

What track is it?

Are you pulling the track from memory or is it a user defined one?

[Evergreen Dan]

Palmer. User defined by walking to s/f and setting the track. Doubt it's a GPS issue, but who knows.

[NotoriousJJE]

Palmer. User defined by walking to s/f and setting the track. Doubt it's a GPS issue, but who knows.

I think this might be your issue. Might not be reading it right. Have you tried defining the track from "inside" the track ie. warmup lap?