Use a video feed of the rider + AI to determine how aero the rider is at all time. Use that information to adjust resistance/speed in game. That will allow us to practice riding faster outdoors.
It’s not April 1st, sir.
1 Like
I’m well aware of that. Hopefully that means the product team will consider it.
You expect every rider to have a video feed? And Zwift to process that in real time? In a large pack? Not gonna happen.
1 Like
I use a M4 Pro Mac Mini and a 55” TV screen on a wall, how should this video feed come from that setup?
How would the video feed of a rider alone on their trainer bike be affected by a large pack?
I’m thinking that the rider’s position would need to be taken into account with relation to their position in the pack in Zwift, as well as the body position of all of the riders around them.
From reading, what the user is looking for is:
I WANT to tweak the aerodynamics of my own avatar in game, based on real-time recording of me riding my own bike on a trainer SO THAT I can quantify the aero benefits and get a better position for riding a real bike outside.
That’s how I understood it. Nice idea, but that’s going to be very difficult to do. I cannot see how it would be accurate.
You have to just do the miles outside and figure out what works. And the most aero position might not necessarily be fastest because you might get uncomfortable and not be able to produce sustainable power.
A good number of rider already have a video enabled device in front of them - I for example run Zwift on a laptop on a stand in front of me - and the laptop only has to work out where I am on the scale between out of the saddle to seated with my nose on the bars (we’re not doing drag down to the nth degree of how flat is my back and the height of my socks). So its not a beyond belief April fools day idea some seem to think it is.
Zwift are already guesstimating a rider CDa from weight and height, and using that in large group draft scenarios. This would just change the CDa values used from static values to dynamic values but its doesn’t add any more maths than a simple multiplier to each riders current CDa.
On the scale of likely in the short term - no - but certainly not a no never it can’t be done thing. We could also chuck in that smart trainer makers are already deeply familiar with strain gauges, and could probably easily design systems to work out a riders probably current position on the bike without even needing a camera.
Like most things with technology its not can it be done - it’s is there sufficient likely benefit to sales versus additional cost to justify funding the development work ongoing feature maintenance.