Hello!
Here goes nothin, This may be a super redundant
But I think fundamentally if the gears start turning on this project and it is refined over time it could become a super useful tool.
The idea behind it is that this “Conebot” becomes open source project for all to utilize / improve to automate track setup & tear down. Without the human need aspect.
(Which kind of has a drawback of lacking that community cone pickup. But I believe most can do without the cone pickup & have done enough of that.)
Who is it for?
People who dont have Access to groomed Indoor Tracks or any sort of Track.
are trying to start a i2s Chapter or pre-existing Smaller i2s Chapters
or even Solo track Sessions
Competitive Racers who have other things to do in the first hour of practice
People with back issues
Ideally someone could just show up early to the shred session and not have to set the entire track up on his / her own. Maybe getting more out of your track sessions with having to worry about setup / tear down. Especially if miraculously your entire turnout does an Irish goodbye on you.
Nobody likes or appreciates that.
I will start developing a chassis as a start soon enough
but I’d like to create hype for this project to gain more input / ideas
Features I’d like it to have.
alternating cone colors / apex turn color coordinating
lightweight and as compact as possible.
the ability to create on the fly track layout
Preset track layouts
quick enough to stay under 1 hour of track setup / tear down.
LOVE this idea for a project. We were talking about somehow mixing a Roomba with a push sweeper to make an automated lot cleaner. If the ConeBot project works out I could see sharing some of the solutions with a SweeperBot.
Ill be sure to make everything available that Ive developed and as much Documentation as Possible.
if Im going to be doing this all on my own it may take a couple years considering some of factors Im already thinking about but Im happy you are interested in it!
I love this idea so much. Now, I have no clue when it comes to ESK8 racing but i can imagine this would be awesome for smaller groups of people getting together to practice without booking time on a track.
If I were you, start basic and add features such as colour alternatives and cone cleanup at a later date. Get the cone dropping mechanism and nagivation nailed down first.
For the cone pickup you’d probably need some kind of camera with object detection and that would become an order of magnitude more difficult to design.
Are there any standardised cone sizes or do you plan to make it more generic?
yeah as far as esk8 racing maybe it if becomes really ironed out it could be of use but initially will be kinda geared towards more for people who want to practice in open lots.
The Cone Pickup bot may need to be a Separate bot from the bot who does the layout
Most of us tend to use the soccer cones but Im sure the design could be altered to pick up different types
Agreed! And picking up the cones by hand isn’t as bad anyways. Placing them out is much more annoying.
I think the main advantage would be that you can automate placing out the same few tracks over and over again, without having to paint the asphalt (which you of course shouldn’t do in a public space) or having to measure out 50 cone locations one by one each time. Also if you paint the asphalt for multiple tracks it gets confusing.
Running the same few tracks over and over really allows you to see how much you are improving, how consistent you are, etc. Having a couple nice layouts and switching between them is great, but not having to come up with new custom layouts or measure out the more complex ones over and over is perfect for repeatability, so you can compare lap times between sessions.
Omni wheels can go really screwy when they lose traction, so there could be a crack in the pavement that rotates the bot 10 degrees, then it loses tracking and the rest of the track loses alignment. So it would need position and rotation feedback from GPS or something.
A bot with 2 driven wheels + front/back casters + encoders might be easier to develop
There’s supposed to be sub centimeter gps positioning now, though I’ve not tried it.
In a previous life I did some work on field positioning of devices - we were using some module. I’ll look it up.
I think it would be a lot more fun to have each cone be its own bot. I actually think the development time would be lower because there are fewer hard problems to solve.
Plus it would be really fun to watch them deploy and retrieve themselves.
Might be fun + easier to build a bot that draws in sidewalk chalk instead of cones, if that’s not a breach of etiquette. No cone dispenser, just extend + retract a spring loaded arm
