ABS Modules/Systems Question

[obnauticus]

Hey,

I think it would be a really good idea to design good breaks and "lock" sensors on each of the wheels so one could theoretically take an ABS system/module from a car and integrate it into the FRC kit/Robot (with the pinout(s) and detailed documentation and basic logic). However, I cannot find any documentation on any modules (they are extremely proprietary) and specifics on "lock" sensors (that's the most detailed description I could find). By moving the ABS calculations to a different system--it should save precious processing power. This way there can be faster response times for image recognition and other things which are also crucial to the real-time calculations that the cRIO may have to do.

I am wondering if anyone here knows of an easy to use and welly documented ABS module. This would enable a team could integrate it into their current control system (with the addition of sensors and breaks). I couldn't find any DIY solutions for ABS modules that other people may have designed for robots on Google.

Thank you in advance

[James Tonthat]

It would probably be easier to write one. The principle isn't hard, just have to gather the sensors. Processing power isn't going to be an issue.

[obnauticus]

Originally Posted by jtkellertx

It would probably be easier to write one. The principle isn't hard, just have to gather the sensors. Processing power isn't going to be an issue.

With the pinouts I COULD find on google, it seemed very feasible to retrofit one into a robot easily. Plus it would be pretty cool for other projects too...

[NoodleKnight]

You'll have a hard time. For starters, the mechanical ABS module is of almost no use to your robot, it was designed specifically for that car and well, no competition legal robot will come close.
Same goes for the electronic module, it's not worth the time and effort to make one work with your robot. It too was specifically designed for a specific car, for that specific car's brake system. It's unique to that car and it's handling characteristics. Moreover the ABS module was designed to control the ABS solenoids to pulse the brakes. I wouldn't advise against putting brakes on your robot. Beyond the extra complexities and extra weight, you can use your electric motor to achieve the same effect.

Like mentioned above, I'd stick to trying to create something in programming. Most modern ABS systems sample individual wheel speed and use that data to determine whether or not a wheel is slipping. Most wheel speed sensors are just hall effect sensors that 'watch' a toothed wheel on the axle or wheel hub.

How you determine wheel lock up can vary, there's probably manufacturers trade secrets in regards to this. Two ways that come to mind are comparing individual wheel speeds, or looking at how fast the wheel speeds have changed. From there you can start making things more interesting and see if they yield nice results.
I've given the idea of an ABS-like system myself and I think that just controlling the increase and decrease of drive throttle is probably as far as I'd go. Beyond that it'll start becoming intrustive for the driver and will only partically fix your slip-n-slide problems.

It's hard to compare cars to robots since even though they have wheels and a motor powering them, they are used differently and move differently. But for what it's worth, in low friction environments the best ABS system won't stop your car from sliding sideways, or in any direction once it gets going. First hand experience, lol.

[Al Skierkiewicz]

Remember that ABS on cars modify the exisiting hydraulic systems which are not legal on first robots. Traction control systems make use of ABS brakes to manipulate the differential into thinking both tires have traction so that one tire doesn't spin while no power is applied to the other wheel.

[Jon Stratis]

We were thinking of trying to implement an abs system this year as well. Has anyone done any tests with the brake/coast headers in a typical drive train this year? What i'm wondering is if the CoF is high enough to get the wheel to spin (in coast mode) when the robot is sliding along the floor. If so, then the simplest ABS system would be to utilize your brake coast header and sensor on your drive train - gear tooth or rotary encoder.

[obnauticus]

Originally Posted by Al Skierkiewicz

Remember that ABS on cars modify the exisiting hydraulic systems which are not legal on first robots. Traction control systems make use of ABS brakes to manipulate the differential into thinking both tires have traction so that one tire doesn't spin while no power is applied to the other wheel.

I wasn't planning on using car breaks. I was hoping to translate the logic from the abs module for solenoids in breaks which I would be designing. You could use software and optical response parts from an old computer mouse for a "lock" sensor.

Anyway...anyone here know of an accurate way of making a tach for each wheel. A photoresistor and black electrical tape is probably not accurate enough. Maybe on the gear assembly?

Thanks

[comphappy]

There are optical encoders included in the kit. You can use these.

[DonRotolo]

You'll want to also read this thread.

[obnauticus]

Originally Posted by Don Rotolo

You'll want to also read this thread.

Thanks a lot. Your post really helps, and you DID use an optical mouse! Nice .