[Stephen Downward] has put together a very impressive Internet controlled robot. There are so many things about his video presentation (also embedded below) which we find delightful. Notably, it’s obvious that he knows what he’s talking about when discussing everything from the electronics chosen for the project, the mechanical assembly and the issues with its current state, as well as the software backend that gives him control of the rover.
The bulk of the rover is the Linksys WRT-54G router which he picked up at a thrift shop. This has been a popular model for building rovers for quite some time. [Stephen] is not driving directly from the router’s serial port, but that could be an adventure for him down the road. For now he’s using an Arduino Mega along with an Ethernet shield to connect the motors to his network. The IP camera on the front gives him the video feed to operate this completely over the Internet using his own program written in C#. He mentions that the CD wheels he has aren’t ideal because of their thin tread area (covered in masking tape) and the inaccurate mounting which leaves one of them at an angle. He’s hoping to design and print his own. He plans rent some time on a 3D printer at the local University when their 3D printing service comes back online.
We think the hardest part with robot building is getting your first platform up and running. Now that he’s got that it’s a matter of making improvements and add-ons. Since he’s got the I/O of the Mega at his disposal we’d like to see him implement a bunch of different sensor: line following, bump sensors, distance sensor, heck… maybe someday he’ll scavenge some Lidar for it!
Good work.
Stuff like this gives me great hope for the future generations. Let’s hope he doesn’t lose that hacker spirit when he grows up. Good job.
Yeah, this next generation is lucky because more parents are getting into this sort of thing I think, I had no support with electronics or any of the stuff I was interested in, mainly because I had nobody that I could relate to, I think that parents are getting a bit more tech savvy now.
I like it, a really fun little project.
Although i think he’s a bit too focussed on 3d printing everything. I guess if he has easy access to a 3D printer then good for him, make the most of it. But if it’s gonna be lots of custom orders that he’s gonna have to wait for… consider other options.
There’s a number of servo wheels already out there in various different shapes and sizes. He might literally be reinventing the wheel on that one. As for making braces and things like that, there’s nothing wrong with cutting some wood you know :p At least for prototyping.
The back swivel wheel was a good idea. Easily adds stability and uses an off the shelf part.
As the author mentioned, with a bit of a hack, he should be able to run the servos directly from the wrt54g and cut out the arduino. But I guess the arduino is an easy testing platform for adding other devices, sensors etc.
Personally I’d use a raspberrypi with a USB wifi, USB webcam and then the servo motors wired up to the IO pins. All of which I’ve done before so I know it’s possible.
I found a 3D-printable wheel on thingiverse – I won’t be designing it myself. The arduino is there so I can add in the future, and I used an ethernet shield because I can easily unhook the router and run it off a wire when I debug.
I tried using a raspberry pi with a webcam, but I found the latency to be too high to actually be worth using.
Thanks for your feedback!
Sounds good, but whats the cost of printing the wheel vs. buying one of the many available ones from ebay/amazon? Do you already have access to a 3d printer or do you have to place an order with a company that offers 3d printing?
Within reason, what you’re doing with the arduino should be possible with the PI (with regards to sensors, servos etc).
I’m surprised you had latency issues with the pi. Part of my home security system is raspberry pis with usb webcams and wifi dongles (connected to my main home network) and I have live footage without any noticeable latency. Since IP cams generally have a 0.5 – 1 second lag, and then the hop through the wrt54g, I’m surprised it’s working out better for you. But if that’s your experience then good on you for testing the difference and finding which performed best.
Printing the wheel will cost less than $1.
I’m getting 50-100ms lag at the moment.
Rather than use CDs for the wheels, peanut butter jar lids (or similar plastic screw-on lids) can make great wheels for servos; most of the time, they already have a small dimple (from the sprue, I would imagine) in the center to align with (alternatively, one can easily find the center using a compass and a few marks). Once mounted, a wide rubber-band stretched over the lip can serve as a simple tire with a fair amount of grip.
Great work Stephen! Make sure to post a new video when you get your printed parts all up and running. I am really impressed with the range of skills you’ve gathered to pull off a project like this! Bravo!
This is a great job! Why don’t you use the routers case as the frame for the wheels? open it up and mount brackets to it for the servos.
Great job! You should be very proud of yourself!
This is awesome. Keep it up.
I was hoping that there was a plunge router involved. Can you imagine what you could do with a router-bot? But I think this project is much safer! Looks impressive.
cannot unsee
he got cat5e for Christmas?
Yup!