Video

The encoders of the motors are used to calculate the angle of the drive wheel and the direction of the module. Because the direction of the module is random, when I power up the robot, I turn the direction until the hall sensor is trigged by the magnet. So the hall sensor is only used for zeroing the direction. After this only the encoders of the motors are used. Maybe I should create a video explaining this in detail.

If you look very closely at the wheel when I start the robot in this video, czcams.com/video/3W0p16vsgqA/video.html , you can see all the wheels changing direction, when they search for the zero position.

It is the Wowstick 1F

End of video should link to demonstrations. If not, then here are a video driving autonomously czcams.com/video/9Od0E53R8PE/video.html, and another driving remote controlled czcams.com/video/90HxqwZaWRA/video.html

It didn't go exactly as planned in the competition, because I made a stupid mistake in the programming. But I did get to use the arm of the robot for solving obstacle with the qr code boxes. It can see it in action here: czcams.com/video/Al4ebXraGM8/video.html

Yes, all the plastic parts are 3d-printet. There is a link to the design files in the description.

Thank you. I'm quite happy of the design myself, and the newest iteration with bigger motors works great.

Nice to meet you there. I hope we will see more teams from Odense next year.

My electronics setup is a bit complicated, because I use it for autonomous driving. If you just want to remote control it, it can be done much simpler, but I don't have much experience with that. I would suggest, that you try to contact others, that have built it. Look at the comments section at Printables here: www.printables.com/model/194299-stair-climbing-rover/comments There is also a link to my custom PCB for the electronics, but again, that is most likely overkill for remote control.

I forgot to mention that I also love your implementation of the magnets for the covers. You rock!

Yeah, the magnets are quite nice. I will share the complete design at some point, but I need to clean it up a bit first.

I'm not fully satisfied with the colors, but had a bunch of the green filament, so ended up using that😀. Still a lot of work to make it actually run.

Have you tried ros2 on your robots?

No, it is just running my own software. I have looked at bit at ros, but it takes some time to understand how it works, and I'm not sure I have the time to invest in that.

Don't get your hopes too high. Currently I'm struggling just getting the cameras to work reliably.

I hope you will get it sorted out! On your old robot the camera seemed to work as intended, but maybe it only had the rgb camera?

Problem was a bad connection in the cable on the RGB camera. I fixed it by soldering the usb cable directly to the camera board. I had the TOF sensor on the blue robot, but never used it because of a driver issue. The sensor was quite new at that time. This year I got the sensor working on a Pi4, but in the end decided to use Pi5 on the robot. And then I had the same problem with missing driver support for Pi5. Arducam released support for Pi5 a few days go, but I didn't have the time to implement it.

Great news! Did you manage to get a run in too? I only do dumb industrial robots which just performs the same actions over and over. At some point I would like to add some kind of vision to them too, but so far I haven't found anything (at a not too crazy price) which can handle what I need.

Yeah, I did run today in the competition, but the result was not great. I made stupid mistake of merging to programs for different parts of the track into one, but never actually tested that it was working. But I did solve the qr code box challenge, so the gripper in the video worked decently. I work with industrial robots and vision in my day time job, so I know that part very well😀

Thank you. It is mostly an evolution of the hardware, but so far it has not run that much, so I'm not sure it is better.

Yeah, it has been a long time since we have seen anyone do a Crazy Ivan

@@wildwillyrobots Maybe next year!

We definitely need someone who can write decent software in the competition😀

These 100mm steps are not a problem, but I don't think it will be vacuuming normal stairs anytime soon.

Slip rings are a bit expensive when you need 6 of them. And also, I'm not sure they can transfer the encoder signals from the motors without adding noise.

@@wildwillyrobots guess you have a point there on cost, idk about the communication, it's been a minute since I seen this.

The bevel gear was modeled in Onshape using a feature script called "Bevel Gear Pair". It was then combined with a normal gear made using the "Spur Gear" feature script.

@@wildwillyrobots Could you possibly make a CAD tutorial on making one? Also if you know is there a way to do this in fusion360?

​@@bio.s2903In Onshape, once you have selected the feature script, you get a dialog, where you put in the dimensions and the gear ratio, and then the gear is automatically generated. I have never used Fusion360, but assume it can run custom scripts to generate complex parts, just like Onshape.

In this video it is remote controlled. I have a Raspberry Pi that receives commands from the remote. It then calculates the required speed and angular direction for each of the 6 wheels, and send those values to a custom build motor driver, that turn does the low level control loop for controlling the motors. This is a somewhat complicated setup, but that is because my primary use of the rover is to drive as an autonomous robot. For someone who just want to drive around with a remote, it can be done much simpler.

What kind of information do you need? There are more details in the link in the description.

Yeah, but still most small robots in the size of a robot vacuum cleaner will not be able to go up.

No 🎮 Fully autonomous🤖

You can find most parts at www.printables.com/model/194299-stair-climbing-rover In the comments at Printables you can also find a link to the motor driver board schematic. I have not shared any code.

I am currently doing a group project about a gripper remote robot thst can climb stairs and turn, and cost of building the whole thing should be lesser than $120(AUD), size is 250 (mm) x 250(mm) . Can anyone provide any suggestions for this one?

Most expensive parts are 6x motors, 6x servos, motor driver board, Raspberry PI, battery, camera. Haven't calculated the exact price, but as it is shown in the video, I think somewhere in the range €250-350.

@@wildwillyrobots wow, that is crazy. Thanks for sharing, despite the price is a little bit out of budget to me, it looks pretty good.

Yeah, it is a bit expensive as you need 6 of everything.

@@wildwillyrobotsyeah... I can get that, I think we may need another plan for this haha

Others have made the same suggestion, but they are a bit expensive. Also, I'm unsure if they can deliver a noise-free signal for the encoders. Regardless, I simply wanted to try something different.

I think it can do about 15 cm without tipping over. If I remember correctly the the ones in the video is 10-12 cm

There is a Raspberry Pi inside, so it is possible. But not that useful for my application.

It is a Wowstick 1F+. It is struggling a bit the first time the bolt cuts the thread in the pastic.

Check out the description at Printables: www.printables.com/model/300105-differential-swerve-drive

You can download the individual parts on Printables. www.printables.com/model/194299-stair-climbing-rover/files

I have mostly used it for autonomous driving and not remote controlled. You can check the comments on Printables. There are some details about the electronics I used.

@@wildwillyrobots hey I have some query can you share your whatsapp number or insta id

It is mostly the design from this video czcams.com/video/TP8wQRZUuk4/video.html (link to model in description). Main difference is that the motors of the robot are horizontal and not vertical as in the video of the swerve module.

@@wildwillyrobots Thank you so much

No, but there is not much to it. Some speed controllers for the wheels and a little trigonometric calculations to find the angles or the servos.

You will need larger wheels to climb normal stairs. This means that you will only have one wheel on each step at a time, which I think will be a problem.

The code is currently a mess, so I haven't shared it. Also it has various dependencies, so it is a bit hard to get it running. But overall I have low level control on a Raspberry Pico, that controls the speed of the wheels and the servos. Then I have a higher level program, running on a Raspberry Pi 4. This makes the image analysis for detecting obstacles, and calculates the required speed and position of the wheels, which is then sent to the Pico.

@@wildwillyrobots is it just a simple feedback from vision that controls the motor velocities or you calculate the position data to drive the robot to the position

@@barknvar3902 Image analysis finds the line in the bottom of the image. Then a P-controller is used to calculate the steering direction. Is is just the error (deviation from center of image) times a constant. This gives a steering direction, and from this the speed and orientation of all wheels are calculated. So it is just a controller trying to keep the line in the center of the image.

Yes, the motors can be positioned freely around the large gear. In this version, I'm also using a bevel gear for the motors, so that they can be horizontal and not vertical as in the prototype.

@@wildwillyrobots Interesting - Thank you - I look forwards to the follow-up video.

Thank you. Actually the control is not that difficult, but more on that in the next video :-)

I neeeeed that video!@@wildwillyrobots

Not sure what you are asking?

@@wildwillyrobots i was there in dtu last month. i did not hear anything about it.

It was in the Library April 12 and 13.

@@wildwillyrobots are you working in DTU?

@@xinglinjiang4952 I'm a former student at DTU