4 ms·

I don't think it's impossible, but to me it's a solution that introduces more problems then it solves. First off, you'd need to either have a a battery to the

I don't think it's impossible, but to me it's a solution that introduces more problems then it solves. First off, you'd need to either have a a battery to the brake lever. If you're going to power it off the main battery, you're going to have to run a wire to then, and if you're routing 2, then you may as well route another 2 for signalling. Otherwise you're throwing out the main advantage of wireless. But that means now you've got an additional battery to monitor. Second is that the comms systems need to have pretty heavy duty filtering and need to do it fast. 100 milliseconds delay in signalling means at 30kph you're going to add almost a meter of stopping distance. That's a lot of processing power and a lot of power consumption to ensure the near microsecond performance that you're going to get with hydraulic systems. Third is the issue of tying your brakes to an electrical system. When your main battery dies, that more or less means that you have no brakes whatsoever. Fourth is that now your components now also have to go through FCC certification, as does any piece of equipment that communicates on radio frequencies. Further increasing cost. And you throw out all the advantages of hydraulic brakes; no power to use, redundancy between front and rear in the event one fails, ease of maintenance for hydraulic brakes, being able to feel the behavior of the brakes under your fingers, etc, easy to notice if there is a problem (leaks primarily will show up as decreasing fluid in the reservoir). Honestly IMO it doesn't really seem worth the tradeoffs.

fouc 5y ago
The brake could be designed to lock the wheel when power runs out. So as power drops, the brake starts gradually slowing the bike.
- oldsecondhand 5y ago
  Thats sounds like a downgrade. With mechanical brakes, you can cycle even if you've run out of power.
- somerandomqaguy 5y ago
  And what happens when you run out of power mid ride? Or you get some failure that causes intermittent cut out (corrosion on the connectors, etc)? You've more or less turned your bike into a 30kg dead weight that happens to have the shape of a bicycle and have to carry it the rest of the way vs just pushing it. Or worse, a bike that brakes on it's own without warning. Like I said, what is the real advantage here trying to go wireless here that makes it worth the cost of loosing advantages of a mechanical or hydraulic system? I could maybe see some advantage of a full brake by wire system over mechanical or hydraulic, but wireless? The weight savings aren't that significant.
robbedpeter 5y ago
Some sort of energy capture loading a supercapacitor to power the brakes might be feasible, with a small battery backup. You don't need a lot of power storage if you can reliably keep the capacitor full during use. An esp32 has way more than enough clockspeed to keep the reaction speed under 10ms. To me, there's a more complex issue, which is the braking feedback mechanism. You can't just have off/on - you need a gradient, and that needs to be communicated to the rider by feel. The same feel has to be perceived across every braking context, or it won't be predictable and people will crash. Any sort of ABS or smart brake algorithm has to predictably augment the human control feedback and seamlessly integrate with motor controls. Using nfmi would be preferable to wifi or bt, to reduce the possibility of interference. Lastly, your system needs to fail gracefully. If the controller or mechanics of a brake fails, the motor controllers have to be aware of it, and limit speed or go into a limp-home mode. So if you have a siloed power store, secure and reliable wireless networking, human compatible controls and brake behaviors, and graceful failure mode, and at least one layer of redundancy, you could have a feasible braking system. Since brakes are almost always conveniently close to the frame, though, in not sure that wireless buys you anything compared to wired. Electronic braking has most of the same problems regardless of wireless, and wireless doesn't simplify or improve any problems (that I can think of? ) Maybe isolating each brake from the drive controller is desirable for reliability? So with all that complexity, using regular old hand brakes seems like the best solution. Even if you want smarter braking, using a system that works with hand brakes is probably better than a wireless or purely electronic overhaul. The sensorial feedback using analog hand brakes is the peak of UI - interfering with it requires some advantage that exceeds the loss of feedback and reliability. I can see wireless controls allowing for novel form factors where brake cabling isn't possible or becomes complicated, but for most regular bikes, it seems unnecessary.
- somerandomqaguy 5y ago
  >Since brakes are almost always conveniently close to the frame, though, in not sure that wireless buys you anything compared to wired. Electronic braking has most of the same problems regardless of wireless, and wireless doesn't simplify or improve any problems (that I can think of? ) Maybe isolating each brake from the drive controller is desirable for reliability? Potentially (and I say this with skepticism) it could simplify the hydraulic lines since if it's full brake by wire then you could only run one system to control the master cylinder rather then the multiple required for ABS. Then you could run something like an armored ethernet cable to the brake lever itself (just using armored ethernet as an example, I've got no idea if it's a good idea or not). Assuming you've got some mechanism to sense what's going on in the hydraulic lines and transmit it back to the to the lever and the hands of the rider, it means that potentially that brake feel can be adjusted on the fly. That said, I don't really view it as a big enough advantage to loose ability to control the brakes in a loss of power situation. >I can see wireless controls allowing for novel form factors where brake cabling isn't possible or becomes complicated, but for most regular bikes, it seems unnecessary. That's really debatable. Mostly because even if you can move your brake lever where ever, it still needs to be within reach of your hands while on the bike's handle bars, and that inherently limits it's placement to where your hands are near. Maybe if you moved the controller to a glove on the hand itself, but then you'd have to think about preventing inadvertent activation on whatever glove you're wearing. And I really doubt anyone's silly enough to think that we should use an app on your smartphone to control your brakes.

fouc 5y ago

The brake could be designed to lock the wheel when power runs out. So as power drops, the brake starts gradually slowing the bike.

oldsecondhand 5y ago
Thats sounds like a downgrade. With mechanical brakes, you can cycle even if you've run out of power.
somerandomqaguy 5y ago
And what happens when you run out of power mid ride? Or you get some failure that causes intermittent cut out (corrosion on the connectors, etc)? You've more or less turned your bike into a 30kg dead weight that happens to have the shape of a bicycle and have to carry it the rest of the way vs just pushing it. Or worse, a bike that brakes on it's own without warning. Like I said, what is the real advantage here trying to go wireless here that makes it worth the cost of loosing advantages of a mechanical or hydraulic system? I could maybe see some advantage of a full brake by wire system over mechanical or hydraulic, but wireless? The weight savings aren't that significant.

robbedpeter 5y ago

Some sort of energy capture loading a supercapacitor to power the brakes might be feasible, with a small battery backup. You don't need a lot of power storage if you can reliably keep the capacitor full during use. An esp32 has way more than enough clockspeed to keep the reaction speed under 10ms. To me, there's a more complex issue, which is the braking feedback mechanism. You can't just have off/on - you need a gradient, and that needs to be communicated to the rider by feel. The same feel has to be perceived across every braking context, or it won't be predictable and people will crash. Any sort of ABS or smart brake algorithm has to predictably augment the human control feedback and seamlessly integrate with motor controls. Using nfmi would be preferable to wifi or bt, to reduce the possibility of interference. Lastly, your system needs to fail gracefully. If the controller or mechanics of a brake fails, the motor controllers have to be aware of it, and limit speed or go into a limp-home mode. So if you have a siloed power store, secure and reliable wireless networking, human compatible controls and brake behaviors, and graceful failure mode, and at least one layer of redundancy, you could have a feasible braking system. Since brakes are almost always conveniently close to the frame, though, in not sure that wireless buys you anything compared to wired. Electronic braking has most of the same problems regardless of wireless, and wireless doesn't simplify or improve any problems (that I can think of? ) Maybe isolating each brake from the drive controller is desirable for reliability? So with all that complexity, using regular old hand brakes seems like the best solution. Even if you want smarter braking, using a system that works with hand brakes is probably better than a wireless or purely electronic overhaul. The sensorial feedback using analog hand brakes is the peak of UI - interfering with it requires some advantage that exceeds the loss of feedback and reliability. I can see wireless controls allowing for novel form factors where brake cabling isn't possible or becomes complicated, but for most regular bikes, it seems unnecessary.

somerandomqaguy 5y ago
>Since brakes are almost always conveniently close to the frame, though, in not sure that wireless buys you anything compared to wired. Electronic braking has most of the same problems regardless of wireless, and wireless doesn't simplify or improve any problems (that I can think of? ) Maybe isolating each brake from the drive controller is desirable for reliability? Potentially (and I say this with skepticism) it could simplify the hydraulic lines since if it's full brake by wire then you could only run one system to control the master cylinder rather then the multiple required for ABS. Then you could run something like an armored ethernet cable to the brake lever itself (just using armored ethernet as an example, I've got no idea if it's a good idea or not). Assuming you've got some mechanism to sense what's going on in the hydraulic lines and transmit it back to the to the lever and the hands of the rider, it means that potentially that brake feel can be adjusted on the fly. That said, I don't really view it as a big enough advantage to loose ability to control the brakes in a loss of power situation. >I can see wireless controls allowing for novel form factors where brake cabling isn't possible or becomes complicated, but for most regular bikes, it seems unnecessary. That's really debatable. Mostly because even if you can move your brake lever where ever, it still needs to be within reach of your hands while on the bike's handle bars, and that inherently limits it's placement to where your hands are near. Maybe if you moved the controller to a glove on the hand itself, but then you'd have to think about preventing inadvertent activation on whatever glove you're wearing. And I really doubt anyone's silly enough to think that we should use an app on your smartphone to control your brakes.