4
$\begingroup$

I was asked this in a phone interview for a robotics job. Googling has not really helped. I assume it is some sort of state prediction model that can be used in a Kalman filter.

Can anyone give me a formal description? A link to a reference would also be nice.

EDIT to clarify, the interview was for a self-driving car company and before the question we had been discussing Kalman filters, Particle filters, and path planning algorithms (A*).

$\endgroup$
  • $\begingroup$ Are you referring to Bike Riding Robot? $\endgroup$ – CroCo Dec 10 '16 at 17:28
  • $\begingroup$ No. The job interview was for a self-driving car company and we had been discussing path planning algorithms, Kalman filters, and Particle filters when he asked the question. $\endgroup$ – somerandomdude Dec 10 '16 at 22:48
3
$\begingroup$

The bicycle model is a simplified model of vehicle dynamics. It's really common so I'm not sure why a web search did not get any results for you.

A random paper, first one in my search that had a diagram and was not behind a paywall...

Modification of Vehicle Handling Characteristics via Steer-by-Wire By Paul Yih, Jihan Ryu, and J. Christian Gerdes.

| improve this answer | |
$\endgroup$
  • $\begingroup$ Isn't the simplified model of a vehicle known as cruise model? $\endgroup$ – CroCo Dec 11 '16 at 15:53
  • $\begingroup$ Sure. One of many. $\endgroup$ – hauptmech Dec 11 '16 at 21:01
  • $\begingroup$ Thank you! Maybe I was using bad keywords but my searches kept on bringing up results related to modelling bicycles. $\endgroup$ – somerandomdude Dec 12 '16 at 2:57
1
$\begingroup$

"The" bicycle model? No idea. But "a" bicycle model is something that's a great jumping point for tons of interesting topics though...

  • holonomics
  • closed-loop systems / feedback models
  • torque calculations
  • gyroscopic stabilization
  • general stability / lyapunov theory

Terrible question for a phone interview!

Bicycle Dynamics and Control by K. J. Åström

This PDF discusses most of those issues.

| improve this answer | |
$\endgroup$
  • 2
    $\begingroup$ I would not say off hand that this is a bad phone interview question. It depends on the type of robotics job and on what the hiring manager was going after. Perhaps the question was phrased to see if you know how to say "I don't know," rather than making something up to sound smart. Perhaps they wanted to see if you could think on your feet. Or if you know Hamiltonian dynamics or nonholonomic system control. Interviewing isn't just about what the candidate knows, but also who the candidate is. $\endgroup$ – SteveO Dec 10 '16 at 16:30
  • 1
    $\begingroup$ @SteveO I did answer I don't know. The thing is I kind of remember seeing something called a bicycle model somewhere used as a state prediction model for a Kalman filter but I can't seem to find anything when I google it. I had been talking about Sebastian Thrun's book and Udacity course and he told me that the model is used extensively in his Udacity course. I can't find anything with that specific name in the course either. $\endgroup$ – somerandomdude Dec 10 '16 at 22:53
  • $\begingroup$ @Hadi see page 11 citeseerx.ist.psu.edu/viewdoc/… :) $\endgroup$ – Ryan Loggerythm Dec 13 '16 at 21:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.