Wireless Brakes: The Future Of Cycling?

Design & DIY
11 followers
10 Video Views·Aug 31, 2026
You can get wireless shifting and wireless dropper seatposts, but no one makes wireless mountain bike brakes. So, I made my own, complete with an app, ABS, and loads of other features. The problem? Well, they're wireless brakes.
AI Generated·May contain errors
Video Summary
AI GeneratedThe creator developed \"Airlock,\" a wireless \"brake-by-wire\" prototype for mountain bikes. By replacing hydraulic lines with electronic actuators and ESP32 microcontrollers, the project enables customizable braking curves, a single-lever system for both wheels, and programmable features like ABS, though it ultimately serves as a conceptual exploration rather than a practical replacement for hydraulics.
These brakes are totally wireless. I can spin the handlebars all the way around.
No fluid, no hoses, no cables, and they feel a lot like normal brakes. Instant, proportional,
and very strong. But nobody sells wireless brakes for mountain bikes. Not yet. We do have wireless
shifting, like I have on this bike, and we have wireless dropper posts, and I fear wireless brakes
are definitely coming next. So I wanted to see what that might be like. Not just the hardware,
but the features they could have, like one lever braking, braking curves, and even ABS.
So I set out to build Airlock, real working wireless brakes from the future. But I started
this project without a clue because I could only find one person who had tried to do this before.
Yes, Blake Sampson built wireless brakes using RC car parts,
including a powerful steering servo to engage them, and it worked.
But it was sketchy. Blake complained of a binary experience where the brakes were either
off or mostly locked up. And while Blake had the skills to contend with this,
the ability to feather or modulate your brakes is important if you don't want to go over your
handlebars. So I decided to use something other than a servo to engage my brakes.
The brakes in your mountain bike will have a lever that's up at the handlebars. Obviously,
you squeeze it with your hand. And then down at the disc, you have a caliper. Now,
this caliper is actually mechanical. And so I'm going to use it in our system. I just need
something to pull it. So let's start there. How are we going to pull this without a hose or a wire
in between? Well, I have to use what I know in my experience is working on cars. And so I thought
a door lock actuator might work. Now, these are inexpensive and quite powerful. You attach a
battery to these two leads here and it'll go out. If you turn the polarity around, it'll come back
in. And it's meant to be operated by hand because in a car you can push the door lock by hand also.
And so when the spring in this caliper pushes back, it can actually push this actuator back
into position. So we only need to make it go one way. That problem solved. Now, there are some
problems with using a car door lock actuator for brakes, but that's for future me to worry about.
Now, back at the lever, we don't just want it to work like a switch and just engage the brakes.
You want to be able to engage your brakes a little bit when you need to just slow down and then a lot
when you really want to lock them up. And so I need something in this lever to make it read
variably. And then the last component is something to transmit the signal between them. And I can
just use little microcontrollers for that. They're off the shelf and we can program them however we
want. First, let's do a proof of concept at the bench. I started with a mechanical disc brake
caliper and some 3D printed hardware to mount the actuator to the bike. Okay, that stops the bike.
Let's see if the actuator can. So I have here a LiPo battery, a switch, and the caliper with a little
cable holder here. And if I flip the switch, it pulls the caliper, not variable. It just
pulls it straight out. So I just want to see if it stops the wheel.
Okay, so if this works, it's the worst it'll ever work because this is far from ideal. So
let's see if it can stop the wheel. We could use a little more braking force and I guess at a more
optimal angle, it will be stronger and we can work on it. There's lots of things we can do to make
this brake harder. And so we have a proof of concept. It was a promising bench test. The best
part about a door lock actuator is you can move it manually. So the instant you let off the brakes,
the spring in the caliper will push it back. But before experimenting any more with the actuator,
I need to design a lever to control it. I started by messing around with stock brake levers,
trying to find a way to read how far it was being pulled. And I decided it would be easier to just
build the lever from scratch using a rotary encoder. Think of a rotary encoder like a digital
volume knob that the lever can turn up or down when you pull or release it. So instead of pulling
a cable or pushing brake oil, my digital lever would poop out zeros and ones. But with a rotary
encoder as the pivot point, the lever felt lighter than anything I had ever pulled in my life. It's
going to feel so weird with the bike stopping pulling this because it is hands down the
lightest lever I've ever felt in my entire life. Like there's just this little spring and it moves
so smoothly. There's no friction anywhere, but I can't see getting tired pulling this. That could
be an actual benefit. The lever would also need to function as the brain of the system. That's
what this is for. A super bright ammo led screen with touch so that I can check battery levels,
change settings, calibrate the lever or run diagnostics. The screen was about $40 with a
built-in microcontroller and even a battery charging circuit. And because these little guys
can easily talk to each other, I could use an ESP 32 in each caliper to receive commands. The
software would need some work, but in the end it would be a way to change settings and adjust how
the brakes actually felt. Airlock was finally taking shape. Airlock combines cutting edge
hardware and software to bring you the most advanced brake lever ever made. With Feather
Touch technology, you're not braking, you're requesting it. That request travels at the speed
of light over our over the bar low latency link and is processed shortly afterwards. No fluid,
no cables, just air. The lever and screen was looking pretty clean mounted up on my stem,
but I wondered how I'd do the same with the caliper or maybe we should call it the drive unit.
Blake had the same problem. While his levers were pretty compact like mine, the calipers looked like
science experiments. And with my door lock actuator, a 3S RC car battery, and a growing
pile of electrical components, I was starting to accept the same fate for Airlock. But I did make
a schematic. An ESP 32 runs the firmware and receives commands. Then the actuator gets its
power through a MOSFET. Think of a MOSFET like a gorilla with a cricket on his shoulder and the
cricket gives him commands. The cricket says punch that guy over there and then the gorilla obeys.
So tiny input, enormous output. The MOSFET is the gorilla, my ESP 32 is the cricket.
So the gorilla drives this, the actuator, the thing that's using all the power. But you can
think of an actuator like a motor and motors punch back. The actuator's coil spits its leftover
energy back out as a voltage spike, which could damage my chip. I know we're getting pretty nerdy
now, but that's what the diode is for. It's a one-way door that catches the counter punch.
If I'm going to be doing this more often, I need one of those big magnifying glass thingies, but
this is basically the thing. Actuator gets connected here. These are the main power inputs
from the little RC car battery. I got a little piece of fish paper between these and I know it's
not pretty, but I don't do this every day. So break number one was starting to look like the
hodgepodge I was afraid of, but would it work? Okay, so if you're new here, what we're looking
for is to plug this in and not have smoke come out, and then we can actually test it after that.
Okay, please no smoke, please no smoke. Okay, no smoke. Our buck converter is on at five volts.
Forgive me, things normally don't work the first time for me.
It's pretty dang instant too. With the electronics seemingly functional,
it was time to test it on the bike.
The 3D printed enclosure for my drive unit was pretty bulky, but this was after all a prototype.
I guess I should put it on the repair stand so I can spin the wheel.
Spin the wheel? We gotta wake it up first? Nothing. What the heck happened? Yeah,
it says the front's not connected. There would be a little green thing there.
I'm actually not getting anything. Ah, still nothing. I gotta bring this back to the bench
and see what the heck is going on. And so it was back to the bench to figure out what came
loose when cramming my electronics into the enclosure. Aha! After poking around with my
multimeter, I found a cold solder joint, fixed it, and put it all back on the bike.
Yeah, now it sort of fits. Okay, are we powered up? Okay, this had better work now.
Might have to wake it up. That was light braking. Now let me go full on braking.
You know what? The release isn't fast enough. Before we start messing around with that,
I'm going to take it out in the driveway and just see what it feels like.
All right, so we're going to see what braking by wire feels like.
Just brake a little bit. So the front was finally working on the bike with the lever,
but it wasn't powerful enough to actually lock up. And this is, after all, called airlock. It
has to lock. So I have to do some good old fashioned messing around. Okay, let's give it a
shot. Okay, we're rolling. Let's do full brake. Okay, it's definitely applying the brakes. It's
not very strong. Try a different actuator. Okay, let's do a little test in the workshop.
Oh yeah, I can brake a little bit. Oh, it didn't stop braking. Oh, just the calibration.
Oh yeah, look at that. I think once we get the angle right, it's going to be really strong.
Another 18 degrees. Freaking wireless crap. All right, be back in business. There we go.
Oh, okay. Let's do a quick test. Okay. Not quite where we want to be, but better than before.
To get the power I really wanted, I needed to upgrade that front caliper,
and leverage, and adjust the angle of the actuator so it wouldn't bind up. But going
back to the drawing board ended up paying off. Okay, so we tried changing the software. We
tried different settings. Today, Airlock is announcing a partnership with Tektro.
We're going to try a better caliper. Whoa, okay. Yeah, I think we're getting somewhere here. It's
time to test this. Let's see if this feels any different. Okay. Quick calibration. Oh yeah.
If I put a rear brake on there, I think we got something going on here. I could have accomplished
this in a smaller package, but it was working well enough for a proof of concept, so it was
time to move on to the rear brake. I had to manufacture that same exact circuit I did for
the front, but this time I was determined to learn from my struggles and make it work better on try
number one. Okay, practice makes perfect. This one is way nicer than the last one, but does it work?
So finally, I had two working brakes and a lever, and I could start working out the final bugs.
Caliper, door lock actuator. This one was a little welding and screw up job over there.
With the Airlock app, stopping is just a sign in away. Notifications, analytics, and account info
right at your fingertips. Braking is no longer tied to your bike. It's tied to your Airlock
ID, so your bike can focus on other things. Download the Airlock app. Get locked in. So my
wireless brakes were pretty much working, but I had a lot of adjusting and tuning to do.
I ended up making a new lever body, a new lever, and changing some code to add features and make
the brakes work better. And now I can give you a full demo and test of Airlock, the world's most
advanced wireless braking system. For starters, Airlock has only one lever that controls both the
front and rear brakes. So I made a feature called braking bias, which lets you control exactly how
much force is applied to each brake. If you move it to the front, your front brake gets more braking
than your rear. But with one lever, we needed some redundancy for safety. So there's a failsafe called
your shoe. Place your shoe on the top of your front tire behind the fork bridge, and your bike
will come to a stop. Now you can see the main screen shows the battery life and your connection
status to the front and rear brake. These little lines next to F and R mean they're connected to
the lever. As you squeeze the lever, it shows your braking percentage in the center of the screen,
just as a number. Now let's say that gets thrown off somehow. You can actually calibrate that.
Just tap the number in the center, pull the brake to the bar or wherever you want 100% to be,
and then release. And it stores the value. And yes, Airlock is designed to be pulled all the way
to the bars. Remember, this is futuristic. And so I made this little tiny lever that actually
clears the rest of my hand. So back to the screen. If you tap this Airlock logo, it enters settings
where you can adjust your braking EQ. This is essentially a six band equalizer that controls
your braking curve. The way mine is set, braking only increases modestly as you pull the lever at
the beginning, and then it ramps up at the end as you pull your lever all the way to the bars.
If you want your brakes to feel grabby, kind of like Shimano brakes, you can make it ramp up
sooner at the beginning by increasing those values. Now, traditional brakes can't truly do
this. As far as I know, Magura's HC3 levers are the only thing that comes close, but most brakes
do let you control free stroke or bite point. That's how far you can pull the lever before the
pads make contact with the rotor. Airlock uses a setting called derp. The higher you set your derp,
the more the system just ignores the beginning of the stroke when you start pulling the lever.
So if you like to ride with your finger kind of pulling the lever a little bit, but not have it
engage the brakes, you can just increase your derp a little bit. But there's another feature
that's only possible with electronics, which does exist on some brakes today, ABS. Now,
not only can you turn anti-lock braking on or off, but you can control the frequency.
When the lever gets to 100%, ABS engages and it pulses your brake at whatever frequency you select.
So 10 hertz is 10 times per second and so on. We'll need to test that on the gravel in a few
minutes, but there are four more features in the screen. Sleep lets you put the lever to sleep,
which does happen automatically and it saves power. Then work mode ignores your input from
the lever in case you're messing around with it or transporting the bike. But then there are two
invisible features. One is firmware updates, which are controlled by me as I refine the lever
software. And then the next one is the log, which records everything that's happening so I can
review it later and chase down issues and refine the system. This can still be improved a lot,
but overall I'm pretty happy with it. It looks clean and it might even resemble what a real
wireless braking system would be. But that clean look ends here. Down at the drive unit,
there's a lot more to improve, but they work. Now I didn't use Bluetooth or Wi-Fi to control
these like you might think, but rather ESP now, which is a super low latency and probably more
reliable protocol that only these ESP 32 units use. That brain is always checking for the calipers
and monitoring the connections so you can err on the side of caution. What do you say we test
these for real? Bring her to a nice stop right in front of Curtis. I don't know why I'm so
astonished that they work. All right, nice smooth stop. You know why I'm so surprised is that a
rotary encoder is not totally smooth. It's got like steps, like clicks. And I thought I would
feel the clicks more, but yeah, it's got enough clicks where you don't actually notice it. And
it's just, it's just crazy. Like this could actually work. The question is, would it actually
be safe? All right. Time to test the ABS. Okay. So first menu system ABS. Okay. It's at 10 Hertz.
Oh yeah. Okay. I'm going to drop in. All right. Full emergency stop.
It definitely makes the bike go further before stopping, but if you were trying to steer,
I guess it would keep being controlled. Maybe I'll try steering a little bit.
Yeah. I mean it's ABS. Okay. Let's do ABS off and go full endo.
Introducing Airlock Max. Unlimited braking every day of the week for just $10 a month.
No more topping up your balance mid-ride. And best of all, Airlock Max includes anti-theft
technology. Steal the bike and the brakes stay behind. They're locked out. You're locked in.
So could we see a product like Airlock and more importantly, should we want one? Well,
there are no confirmed wireless brakes in development as far as I know,
but it wouldn't surprise me. And for people with hand issues, adaptive riders, or even
a remote brake for a bike trailer, this could be legitimately useful. And I personally like
the ability to slow my kid down remotely. Not to mention those other features like
braking curves and ABS that I didn't even think of before I started working on this.
But the truth is most of us don't want wireless brakes. Wireless brakes would be perhaps the best
example of a solution looking for a problem. And the potential liabilities should be obvious to
anyone. Hydraulic brakes are really, really good. This bike is a perfect example. I chose this bike
specifically because it's the bike I did my last brake test on. Instead of Shimano brake oil,
I bled these brakes with baby oil. That was seven years ago. The baby oil is still in the brakes. I
took off this bike and they still work fine. Hydraulic braking systems like the one this
bike originally came with have truly stood the test of time. And people are getting really tired
of forced innovation because you sometimes get forced into it. But I just don't see us getting
forced into wireless brakes. So I'm actually looking forward to watching it happen. And I
think if it does happen, nobody will be surprised. When I sent my friends screenshots of my fake
Airlock website, they all gave it a double take. They thought it was real until they scrolled down
and saw how silly the caliper looked. So if you want to play a joke on your friends, you can visit
my fictitious Airlock site in the footer of bermpeak.com. I'll leave it up for a month or so.
And as for my brakes, Airlock is introducing a new feature today. Airlock is proud to offer a
CoasterLock. If Airlock fails or falls asleep for no reason, just kick your foot back and the rear
wheel will brake, stopping your bike wirelessly. With CoasterLock, you can err on the side of
caution. Airlock. Get locked in.
