Censored Skateboard

Details

Thought about it for a while but just started. not sure if this fits any category for now...

I can't do any tricks beyond jumping onto my board but not the skater way...

But I know enough tricks from watch xgames.

The sensors will mount to the inner bolts of the back truck, with the distance sensor facing down.
This gives a direct Distance to ground measurement to add to the imu data.
Most skate tricks consist of rotating the board in one or more axis.
combined with the pico's RTC(Real time Clock) which is pretty good! should be able to get extremely precise data
App Feature - use the Phone IMU to "record a trick".
- Then match this trick to sensor data
- note Dominate axis/secondary axis (what counts)
Base electronics weight: 20g
+ 15g AA battery
+ Case
Total Weight about 40-60g Compared to a total Board weight of About 2500-4000 g = 2% weight

I mean what Robot would do this? not my video, lands it at 4:19

Basic plan:

Dumb Sensor on Board.
Smart App wifi/bt connected.
Machine Learning to help analyze data and spot "Tricks".

Its nice when things work! Day one.

After a few minutes of the wifi being annoying and then android not wanting unencrypted urls.

Got the android app to load sensor data (111 reading, at the time of loading)
update pico time from using android as an http argument url (10.0.0.11:port/CurrentTime)
Next add timestamps to the imu data/
- log data evenly 50x/100x per sec
sync data on android side and start plotting/manage the data!
3D Print a full case + Battery. (i think just 2x aa?)
check picoW power use with wifi updates about every second
- Check pico update speed
- eventually try bt
- maybe even micro python if that works first?

Project Logs

Collapse

Virtual Board.
Saul • 05/09/2023 at 07:25 • 0 comments

Took some scissor hand work with the code to get it working in my app but really not bad at all.
Now I just have to figure out how to rotate the board vs the "camera".
The sensor connection really needs to be ble for an easy setup.
not my first board. + UDP
Saul • 05/05/2023 at 05:07 • 0 comments
I think ble makes more sense for this, better power use and no need for a common wifi connection.
But for now I got a better wifi connection. After going through most of the wifi example folders I was about to switch to python when I found a udpclient example that works with the pico.
Much faster connection than loading a webpage.
- anyways needed a model to mirror the sensor.
- I think the wheelbite cuts can be a bit more smooth.
- and the nose/tail should be round so they don't hook on something i guess...
I haven't been able to get the mpu to work with the esp8266 nodemcu. but this circuit would nice and tiny.
I hate it.
Saul • 05/02/2023 at 05:57 • 0 comments
I think I'll ditch the distance sensor for v2.
1. its just annoying how much space it takes.
2. with the case its 2x the weight i estimated.
maybe i'll wire it in as an addon. with its own case..
the case style should be good for something that doesn't need to be as portable...
Basic prototype.
Saul • 05/01/2023 at 16:24 • 0 comments
V1 case. 2aa batteries direct.
- 44x80x30mm (38 with sensors) 94grams...case has 4 loops 1.6mm walls.
- power use was about 40 idle and 80 peak on updates. so about 30 hours per pair...
- with about 10 ints per data point plus the date, should be able to store about 1.5 minutes of data at 50hz.
- if i can use program space its more like 15minutes of data. but send out 2mB bursts might need a better connection.
tested this stl viewer demo app. will probably try to add this so I can create a Digital Board that mirrors the sensor data.
Then plotting some data might be useful once it has context.
https://github.com/dbrant/ModelViewer3D