Blog · · 5 min read
Will my hardware idea work? An AI feasibility check on a spoke-magnet bike light
A spoke-magnet bike light sounds simple. A quick physics check showed why a plain coil would barely light the LED, and how three concept options fix it before anyone prints a part.
Every hardware project starts with a sentence. Here it was this one: put magnets in the spokes close to the hub, mount a coil on the fork, and let the spinning magnets light a red rear LED and a brighter white front LED. The housings should be 3D printed and clamped on with screws so nobody can steal them. Two-wire cable, zip ties, done.
It sounds like an afternoon project, and parts of it are. But one question decides whether it is worth printing anything: will the coil produce enough voltage to light the LEDs at normal cycling speed? This article follows how the concept step on makemachine.ai answered that question for the public project Bike Light, and why the honest answer was "yes, but not the way you drew it".
The question an idea does not ask
An idea describes what you want. It rarely describes the numbers that make it work. Before CAD, before a parts list, a useful concept step turns the sentence into a few physical checks:
- How fast do the magnets pass the coil?
- How much voltage does that induce?
- What does the light actually need to switch on?
- What happens when you stop at a traffic light?
None of this needs a lab. It needs the right formulas, sensible assumptions, and the discipline to write the assumptions down. That is a job AI does well, as long as it shows its work.
Why "close to the hub" is the hard part
Induction depends on how fast the magnetic field changes in the coil. A magnet that sweeps past quickly induces a higher voltage than one that drifts past slowly.
On a wheel, every point turns at the same rate, but not at the same speed. A point on the rim moves as fast as the bike. A point close to the hub moves only a fraction of that, in proportion to its distance from the axle. Putting the magnets near the hub keeps them small, safe and out of the way, and it also makes them slow.
A white LED does not glow gradually from zero. It needs roughly three volts before it lights at all, and more before it is bright enough to be useful. A slow magnet and a plain air coil add up to a voltage that stays below that threshold for most of a normal ride.
The concept's model put a number on it: with a plain coil, the white LED would only start to glow above about 30 km/h. That is an estimate from a model, not a measurement, and the project says so. Still, the conclusion is clear enough to act on. A light that only works when you race downhill is not a bike light.
The fix: give the field a path
The concept's answer was not to give up on the idea, but to keep what mattered to its author and change one part. The magnets stay near the hub. The coil stays on the fork. The two-wire cable stays. What changes is the pickup: it needs an iron core.
An iron core guides the magnetic field through the coil instead of letting most of it spread into the air. The same slow magnets then produce a much stronger change of flux in the winding, and the voltage rises with it. This is the same reason commercial dynamos and transformers are built around iron, not around air.
Three options instead of one answer
A good concept does not stop at "add iron". It shows the trade-offs, so the person who has the idea can choose. The Bike Light concept laid out three options:
- Clip magnets. Single magnets clipped to the spokes. The simplest and cheapest build, but each magnet gives one short pulse, so the light flashes faintly instead of shining.
- Magnet ring. A ring of magnets near the hub gives a continuous stream of pulses. The estimate: about 20 lumens at 15 km/h and 35 to 40 lumens from 20 km/h, steady and without a battery.
- Ring plus battery. The same ring charges a small storage cell, so the light holds a constant output of about 50 lumens, also when the bike stands still.
The recommendation was the magnet ring. It is the only option with a steady, useful light that needs no battery, which keeps the build close to the original idea. The battery version solves the standstill problem, at the price of one more part that ages and needs charging logic. That choice belongs to the person building it, which is why it is written as options and not as a verdict.
The rules question
The idea also asked a question that many makers have: Germany has rules for bicycle lights, but do they apply to a light you build yourself, or only to lights that are sold?
Lights for road use in Germany fall under the vehicle regulations (StVZO), and approved bicycle lights carry a test mark, a wavy line followed by a K and a number. Whether and how a home-built light can be used on public roads is a legal question, not an engineering one, so the concept flags it instead of guessing. The next step in the project, research, is planned to look at it with sources. Until then, the honest status is "open".
What AI does well here, and what it does not
This kind of check plays to the strengths of AI:
- It turns a sentence into the few numbers that decide feasibility.
- It knows the standard physics and applies it consistently.
- It keeps the author's constraints (magnets near the hub, coil on the fork, simple cable) while it looks for a fix.
- It writes down every assumption, so anyone can challenge the model.
What it cannot do is replace the bench. The lumen values are model estimates. The real output depends on the magnet grade, the air gap, the number of turns and how well a printed bracket holds the coil steady on a bumpy road. The first prototype will be the real test, and the concept is built so that the prototype can answer exactly those open points.
The value is in the order of things. Finding out that a plain coil only lights up above 30 km/h costs a few minutes in a model. Finding it out after printing housings, winding coils and mounting everything costs a weekend and a drawer full of parts.
Follow the project
Bike Light is an open hardware project on makemachine.ai. The concept with its three options and an animation of how they work is on the project page; research and a design study are planned next. For the bigger picture, read what AI can and cannot do when you build a hardware product, or how the 3D model is made in CAD as code.
Watch hardware get built by AI. Follow public projects on makemachine.ai, read every result in full and get an email when there is something new.
More from the blog
- Will my electronics enclosure overheat? A five-minute heat estimate for a sealed box
- Joining 3D printed parts: screws, heat-set inserts, snap fits or glue?
- Free or paid AI 3D modeling? What you really pay for, and a 30-minute test to compare tools
Open-hardware projects built with AI
- DIY Automatic Plant Watering System: An open-hardware watering system for indoor plants: a small pump, a refillable tank and a soil moisture sensor in each pot keep up to six…
- DIY BBQ Temperature Controller: A temperature controller for charcoal grills and smokers that makers can build themselves. A pit probe, two meat probes and a small fan on…
- ZählerSchau: ZählerSchau is a small, battery-powered unit that clips onto an electricity meter in Germany (water meters via a clamp adapter), takes a…
- Roadstoy: Roadstoy is a remote-controlled toy version of the new Tesla Roadster that can briefly jump and flip using compressed air. It drives on an…
- The forever Washing machine: The forever washing machine is a washing machine offered as a service instead of a product. Customers do not buy the machine but pay for…
- Bike Light: I want to build a bike light that needs no battery. Magnets sit in the spokes close to the hub, and a coil on the fork picks up their…
- Rooficle: I want to develop a light, efficient and affordable e-bike that can be used legally on public roads in Singapore. Its core is an…
- BBQ Modular Grill: I want to design a modular grill. Instead of one sealed block, it is made of separate parts that can each be bought on their own online…
- Go FANUC Go: A mobile base that moves an existing FANUC M-710iC industrial robot (about 560 kg, with a milling spindle at the wrist) out of a room and…
- Jump & Run Robot: I want to build a small, low-cost toy humanoid robot that people can put together themselves. It should be about 30 cm tall, or as small…