Let's Make Robots!

"Pedometer" for roller skates?

So as I mentioned a few months ago, I'm currently training to join a local roller derby team. I am, for lack of a better word, quite out of shape. I can only afford to go skating once a week, and want to make sure I am getting the most of my sessions.

I've seen pedometers for runners, and even small GPS units for skaters going outside, but what if you're inside?

I'm a quad skate snob, and will use my indoor skates for INDOOR ONLY. However, I really want a way to track how far I've gone during my sessions. Even the knowledge that I'm going a wee bit farther every time would be exciting. I get disheartened easily and want something to keep me going.

I'll first post pictures of my skates to get creative juices flowing, then spout off my own ideas:



OK. So I'm thinking one idea would be to paint on the inside of one wheel and use an encoder, but I'm not sure how high I need the proc to count.

The top two pictures show where ideal proc placement would be between the shoe and plate. I want it in a place that is easy to get boots on, and won't hurt me if I fall, because I WILL fall.

Waterproofing isn't a big deal, but if it works out well, I'd like to eventually make a waterproof version for when I can afford a second pair for outdoor use.


Any thoughts guys?

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
I haven't made use of one yet, maybe someone else has. what about an eccelerometer?

did you mean an accelrometer?

that might work  but i wouldnt know how to calculate distance without calculating speed  but it might be able to detect when your foot goes up or down 

why not look into the bicycle meters that do distance it might give you some better ideas i think they use a roller that has an encoder on it, if you add a roller  put enough tension on the wheel so it doesnt roll  when  lifted off the floor

I've actually been researching this. There's a very long equation to figure out velocity using an accelerometer. I've been trying to find websites that go more in-depth, but am coming up empty handed.

It might make everything TONS easier if I forgo putting something on the skat and just clip an accelerometer to my shirt. That way there would be much less calculations with my foot going back and forth, and if I put it on my shoulder on the shirt-collar, it would be out of the way if I fell. AND I could use it for biking/walking/etc.

Using acceleration is an interesting idea. I think you can use the euation for calculating velocity from constant acceleration, v=v0+at, where v is velocity, v0 is original velocity, a is acceleration and t is time. If you sample the acceleration from your accelerometer every second, you can "pretend", that the acceleration is constant for that second until the next sample. Of course this is inaccurate, but you can increase the accuracy by increasing the sample rate.

Don't know if this makes sense or is at all correct.

Well, with sampling, I've always been told to sample at double the speed. Looking at how long strides are on skates as my base, I'd want to sample 2-3 times per second, which would imho be accurate enough for me.

I also need to figure out storing things in eeprom so that if the batteries die I still have my data. Ideally I'd like to take the distance in each minute, and store it. That way I can tell when my fastest minute is, average speed, etc.

2-3 times per second should be enough, if you put the accelerometer on yourself, because your body wont be accelerating as much, but you will be moving the skates back and forth a lot, so they will have a lot of accelerating and decelerating and may need more samples.

I think the final design will use accelerometers. The only thing I have to figure out now is an easy way to get all the information saved on the chip to either a text file or into an excel sheet.

I really wish I had paid more attention to computer programming instead of just hardware...

Some of the PIC18 chips have USB support, so if you use that, you should be able to hook it up to your PC and download the data. The data can be stored on a I2C eeprom. On the PC, you should have a program that can read the data and save them to a file. Read all values (time, acceleration, speed, position etc), separated by a ','. Each line should be separated by a CR/LF. This file, you should be able to import into Excel.
How about a couple of neodymium magnets fixed to the wheel somehow and a hall-effect sensor to detect them?  I'm suggesting two magnets, or maybe four, to keep the wheel in balance.
I'm not too sure about magnets. First there's not much room between the wheels and the trucks. I also don't want any to  possibly fly off. It only takes a very small pebble to catch a wheel and cause a faceplant into the floor (I did this when they didn't clean all the cake off the floor from a birthday party).