My first robot.

I am also following Paul McWorters class. I tried to purchase the sunfounder hat but they were out of stock so I am building the hardware on a bread board. That means i can't use the sunfounder library since it written for the sunfounder hat. Hopefully I can get over the hurdles of everything working together and no dependency issues with the software. Didn't mean to jump the thread on the robot cars but wanted to let you know if we get stuck on something we can help each other.

@noweare

I'm a few lessons behind in that class. I purchased just the board (not the whole kit since I have most of the other stuff). I keep getting sidetracked, spending too much time on other stuff. As I mentioned in the chat, I also bought the RPi AI Hat and have spent 'way too much' time trying to get that thing to work! Hope-fully, I'll get back on track with Paul's course and get caught up. I'd like to (someday) use voice to control my robot car! Right now, I'm trying to get a PS4 controller to work with it using an ESP32 as the comms interface. Haven't quite figured that out yet. 

Regards,  Lee

@lee-g

As my bot ventures thus far have used 4 wheeled bots where one manoeuvres the bot by skid steering  I was thinking of getting a front wheel steering bot to test out the programming required in that scenario to keep it moving along the desired path.   On further looking at the sunfounder bot I mentioned previously it seems to be sort of front half of the car steering rather like my zero turn lawn mower rather than proper front wheel steering if you see what I mean,  so I've abandoned that idea until I find a more suitable chassis to play with.

I expect thinking out how to use those mecanum wheels on your bot to autonomously make it navigate a desired route will be an interesting challenge😀. I see why you are gearing up to utilise AI control on your bot.   

I will probably have a look at those videos too and see how to utilise AI augmented control of a bot.  (and I've just had a peep and see rather a lot of episodes).  To keep the battery on the bot from being drained too quickly with the AI demands I think I will have to consider how much to actually put on the bot though.

Well its an interest challenge you have undertaken and I expect you will learn a lot.  I also prefer using Python or Micropython for developing stuff.  No more compiling, type and run!  

Enjoy the Paul McWorters classes and I will probably follow along in your (and @noweare) footsteps.  Time to swot up 😎 

@byron Did you look at the yahboom ackerman chassis?

@huckohio 

Thanks for the idea of the yahboom ackerman chassis.  I had a look at their web site, but having an old house, mainly carpeted, but some places with floor tiles with grouting groves,  I think the small wheels may  get the bot thrown about a bit.   I'll think I will look to see if I can find a small RC ackerman steering car with bigger wheels that I can adapt. 

I don't want to spend too much on a small indoor bot as its just for a test to see how well I can make it steer to follow a course with front wheel steering, eventually to use the technique on a larger outdoor bot.  The bot I used in the past had skid steering and I think it will be easier to keep the bot heading to its target with a front wheel steering bot. 

To keep this bot on track I expect to use a servo motor controlling the steering and a magnetometer to check the heading. I have found magnetometer readings a bit unstable indoors, on my desk anyway.  In programming these magnetometer sensors they are started with a calibration sequence and I've read that the calibration settings can be different even from desk to floor height.  I wonder if this means a calibration should be redone as the bot moves from room to room.  I don't think I gave this a proper test in the past so I plan to have another go, and this time I cant blame meandering on the skid steering.  😎

@byron 

Have you ever consider those big radio controlled toy cars and trucks. I don't think you will need servo motors to steer them just what is already there which is often just steering left/right/straight. Some have big wheels and some are very large ideal for inside or outside.

https://www.instructables.com/RC-Car-to-Robot/

Posted by: @byron

↑

I've read that the calibration settings can be different even from desk to floor height. 

I hadn't thought about that. That would be a worthwhile test. Does the tolerance of the steady-state pitch-yaw-roll measurement vary with height?

Posted by: @robotbuilder

↑

Have you ever consider those big radio controlled toy cars and trucks.

Yes, I'm going to base my next outdoor bot on a big RC car chassis of some sort, and I think I've found a smaller one similar to your link, which seems it might do for an indoor bot.  Most of these RC cars seem to boast on how fast they go whereas I want one that trundles along.  I think I may have found some that are called rock crawlers which I hope are suitably geared for house creeping and are small enough for this purpose.  Thanks for the link and suggestion.

Posted by: @tfmccarthy

↑

I hadn't thought about that. That would be a worthwhile test. Does the tolerance of the steady-state pitch-yaw-roll measurement vary with height?

I'm not sure about that, I will have to get the magnetometer sensor into a test rig and do some proper testing.

I do remember when I was having a go a couple of years ago that the compass readings I was getting would vary considerably depending on the tilt of the sensor board.  i.e one could get a compass reading say of North, but keeping the sensor stationery apart from just tilting one side up the compass could then read East.  It may well have been errors in my program though.  I will have to start again to check it out.  I'll be interested to see  how your bot's MPU readings compare if you get to check it out.

Posted by: @byron

↑

I'll be interested to see  how your bot's MPU readings compare if you get to check it out.

sigh. It looks like I may need to do extensive testing. I've run into a roadblock with the SmartCar v4.0. After successfully running the tutorial demos (with one exception) I tried to run the complete sketch.

It failed.

Of course, there's absolutely no feedback from the system as it's using practically every byte just to run. So, I have to develop an extensive unit test process to diagnose the issue. The MPU testing will be part of that so I might as well do this test as well.

This was my summer project but isn't how I expected it to go.

@lee-g 

I started this new thread, so as not to 'steal' the other thread.

Only to have it 'stolen' as well.

That you want to engineer your own working project rather than simply assemble someone else's project (or kit) is for me what makes it worthwhile and a real learning experience no matter how far you go with it.

👍

Posted by: @robotbuilder

↑

I started this new thread, so as not to 'steal' the other thread.

Only to have it 'stolen' as well

Are you accusing me of "stealing" this thread?

@robotbuilder That's OK, you guys are taking about things that interest me, so I'm learning as well. I've been working on my car project for 'a while'. I keep getting sidetracked and now that I'm following McWhorter's course again, time seems to slip away. I have the car moving with the Pico (using an IR remote). I'm trying to get the comms figured out to use a Playstation 4 controller as the controller for the car. McWhorter's course has generated a new idea... voice control. So, we'll see how it goes.

@tfmccarthy Who me??? Nah... I wouldn't do that!  😆 

Posted by: @lee-g

↑

Who me??? Nah... I wouldn't do that!  

Oh yes, you would!

You'd do it in a heartbeat.

Twice on Sunday.

😜