Why not automatic transmission?

[Duxuk]

With electrically powered shifting like Shimano Di2 you could have completely automatic transmission. Imagine coasting to the bottom of a technical climb and always being in the right gear when you start to pedal. Or rolling a downhill and never needing to think about changing gear?
Freewheels at the bottom bracket have already been tried so that you can shift whilst coasting, rather than only when pedaling. A fixed cassette at the back and all that's missing is the computer control. You could have a manual control for your chosen cadence and you're in business.
Better get working on my patent application.

 

[tintin40]

sounds like a nightmare to me. Manual gear changing works fine. And if any problems then easy to fix out in the field to get home on. All these new ideas are just making cycling more complicated and less fun.

 

[CTK]

Would be good for beginner cyclists I think, but the cost prohibitive for beginner bikes.

 

[legrandefromage]

Sooooooo 2016...

[youtube]https://www.youtube.com/watch?v=knMPK3TziPs[/youtube]

https://www.cyclingweekly.com/news/prod ... ing-288890

 

[legrandefromage]

Back in 2007

https://www.bikeradar.com/features/revi ... -coasting/

 

[velomaniac]

Re:

Those Autoshift mechs were clever in theory but bloody awful in use. Having next to no control over shifting meant 90% of the time you felt or knew with certainty you were in the wrong gear !
Yes I had one, I just had to give one a go but never got on with it. Aparently you could with practice get the autoshifts to happen under a degree of control by varying wheel speed by using your brakes and reading the terrain.
When you are the engine and you dont have inexhaustable reserves of power like a car engine then having no input to gear selection is painful !!!

 

[klunkrider]

Re:

My dad has just bought an e-bike with Alfine di2 and it has an automatic shift feature.

It’s pretty horrible to use.

It chooses when to shift based on cadence so you are always riding at someone else’s idea of the ideal cadence. I think it can be tuned but it’s also pretty sensitive so it’ll detect your pedalling too fast, change up a gear and then detect you’re pedalling to slow so it jumps around a lot.

Could probably be programmed a bit better but there’s always going to be some sort of quirk that doesn’t work for some riders.

 

[mattr]

Because for anything other than gentle commuting it doesn't work without additional sensors.
Cadence isn't the only thing you need to measure, and you also have a wide range of "good" cadences. (typically 30-40 rpm comfortable range for most cyclists.). You'd also need to look at pedal force, if the rider is stood up or not and the gradient (riders tend to self select a lower cadence with higher force to develop the same power on climbs.) All of which affect the ratio that would be chosen.

Could be done, but really wouldn't be cheap or easy.
Much like an auto box in a car.

 

[jimo746]

Re:

It could be done, with a bit more R&D and more £££ thrown at it.
But is it needed? (mind you that ne er stopped the various wheel sizes emerging!).
Maybe with e-bike popularity it'll come around sooner, but ultimately it'll have to be commercially viable /profitable for it to go ahead in a big way I suspect.
There's probably a man in a shed somewhere in the UK with a very efficient fully working auto transmission, just waiting for his time on Dragons Den

 

[Duxuk]

I expected someone would have tried this sort of thing. I only posted to see what others thought. Personally I like to keep it simple and think that any electrical power, no matter how little, destroys the purity of a bike. Power should only come from the legs IMO. The UCI seem to dislike progress. They banned the Moultons when they broke records. They won't allow recumbents and streamlining isn't allowed. Do they allow Di2 in racing and if so why?