A patent for detecting wheel spin

This one began in a meeting room, presenting a concept to my mentor at Volkswagen.

The problem was slip detection: working out that a wheel has stopped gripping and started spinning. Every traction control system has to do it, and the established suppliers had long since solved it well enough. What I had was a different way of noticing, and the difference was not accuracy — it was speed. About ten times faster than the approaches then available from the major Tier 1 suppliers.

My mentor’s response to the presentation was not “good work”. It was that we should file a patent on it.

Why speed is the whole argument

A traction control system cannot act on what it does not yet know. Everything downstream — cutting torque, redistributing it, stabilising the vehicle — waits on the moment of detection, and until that moment arrives the wheel is spinning and the car is doing something you did not ask it to do.

That mattered particularly on the architecture I was working on. A plug-in hybrid with an electric rear axle has an actuator that can change its torque almost immediately — far faster than an engine can. Give a system like that a slow detector and you have paid for a fast actuator you cannot use. The bottleneck moves from what the hardware can do to when the software finds out it needs to.

Which is a lesson I have met repeatedly since, in places with no wheels in them at all: the speed of a control loop is set by its slowest link, and the slowest link is usually detection rather than actuation.

Filing it as a student

I was on a bachelor thesis contract. Nobody had told me that the output of a thesis might be an application to the German Patent and Trade Mark Office, and going through that process — the drafting, the claims, the specificity a patent attorney requires of a sentence you thought was already precise — was an education by itself.

Writing a claim is not writing a description. A description explains what you built; a claim defines the boundary of what you are asserting, and every word in it is load-bearing in a way engineering prose usually is not. It was the first time I had to state an idea to a standard where ambiguity was a defect rather than a stylistic weakness.

That habit came with me. Requirements are claims too.

The application was filed in July 2020 and published as DE 10 2020 208 741 A1.