Welcome to my newest adventure, the design, construction, and testing of a robot that can go beyond my house and explore the great outdoors!
This is a collection of documents, some of them short, some of them long, where I will describe every phase of the construction of the robot.
“Design Requirements” is an official-sounding way of phrasing the simple question “why am I building this robot?”. No matter how it is phrased, the response is the same.
The robots that I usually work with are great at moving around on carpets. Or at least some of them are. Those 15-dollar “robot car” bases that I’m apt to use have difficulty with high-pile carpeting or throw rugs.
None of these designs could make it very far outside. While they might be able to maneuver reasonably well on a sidewalk or paved surface, they would be hopeless on the grass or on a dirt or gravel path.
So they wouldn’t be very useful in prototyping any outdoor robotics designs. And there are a lot of uses for outdoor robots.
- An automated assistant for your gardening and lawn maintenance tasks
- A security device to patrol your property
- Something to startle your neighbors’ dog.
While I have to confess that my needs probably fall into the last category, it’s not all about making something to give the pooch next door a shock. It’s also about getting over some of the design requirements that are unique to making an outdoor robot.
An outdoor robot, no matter what your reason for building it, has some design requirements that are indeed unique:
- It needs to be able to maneuver on a variety of surfaces, preferably without damaging or defacing those surfaces.
- It should be able to run on its own power for a reasonable amount of time. The term “reasonable” is subject to your own definition (I’m aiming for about an hour minimum).
- It needs to be able to survive an occasional rainfall. Not necessarily a monsoon, but the kind of rain that I, as an indoor-loving human, could be expected to withstand without undue complaining.
- It must withstand ambient temperatures as low as 0 to as high as 40 degrees Celsius. While I do live in a climate that gets much colder than that, I have to respect that I’ll be powering this beast with LiPo’s, and they don’t like operating below zero. If I were building a “snowbot” then I’d look at other battery technologies. On the upper end of the temperature scale, it needs to be able to bake in the sun for a couple of hours without overheating.
- It can take advantage of GPS but will require something more accurate for navigation. Did anyone say “GPS RTK”?
Many of those requirements, such as surviving rainfall and not overheating, will be a challenge to pull off simultaneously.
But that’s where the fun of it is!
I’m building the robot around a 6-wheel chassis, mostly because I can buy one premade and I would rather concentrate on the electronics and programming and keep the mechanics simple.
Having said that, I somewhat regret it.
If I had the mechanical aptitude then I would definitely gather together six motors and put together a chassis with a suspension. But as I’m rather inept mechanically I decided t purchase one, and in my price range, there were not a lot of options.
While the chassis I chose does have a number of merits, it also has a few severe deficiencies. I’ll explain those in detail in another article.
But it isn’t all doom and gloom, and I’ve managed to construct my robot around this chassis despite its shortcomings. And the chassis does have a nice suspension and big fat wheels, so it should be good for outdoor use.
And just to get it out of the way at the start of the project, when I say “six-wheels” I mean six powered wheels, powered in two sets of three. So I’m not controlling the individual wheels, at least not right now. Six motors, two groups of three in parallel.
Lots of Technologies
One other reason for building its robot is the number of technologies that will be required to complete the task. Many of these technologies are ones that we haven’t, as yet, covered here in the DroneBot Workshop.
So now we have an excuse to do exactly that!
You’ll be seeing many of my videos and articles gravitating to technologies used in this project. We have already seen this, with the introduction to RC remote controls done earlier this year, as well as the episodes on using GPS and using waterproof ultrasonic sensors.
Here are a few other topics you can expect to see soon:
- LiPo Batteries – These provide power for the rover, and they need to be handled with care.
- Electronic Speed Controls – ESCs are used to drive the motors on the robot.
- NVIDIA Jetson Nano Stereo Camera – For experimenting with depth perception.
- GPS RTK – Real-Time Kinetics is a method of getting better resolution with GPS.
- Pixhawk Flight Controller – The “heart” of this system, a power dedicated microcontroller.
- ArduPilot – An open-source navigation system that will let us have complete control over our robot, as well as allow us to plan missions for it to autonomously navigate.
- Telemetry Radio – Sending small bits of data wirelessly, with impressive distance.
- Outdoor Sensors – Some special sensors we will use, made to survive outdoors.
- ROS – The Robot Operating System.
As well as all that, you’ll also see some of the construction techniques I employed to put it all together.
In order to assist those of you whole are foolhardy enough to want to build your own rovers, I’ll be providing complete parts lists with every section. And I’ll try and make as many videos and take as many pictures as required so that you can see how all the fiddly bits go together.
I’ll provide wiring diagrams and code samples as we progress.
As the project gets moving, I’ll release more articles and videos. Be on the lookout here, as I’ll likely release many articles without videos.
I’ll also be assembling some more “formal” documentation, details on that will be released when it is ready.
And I’m going to post a number of updates, including video clips, on the DroneBot Workshop Forums.
If you want to keep up to date, the best way is to subscribe to the newsletter. When I released new content, I’ll always mention it in the newsletter.
So, with all that said, let’s get to work and build a 6-wheel rover to scare the neighbors’ dog!
testing – last try got an error message….
Hi Bill, the reason I just sent a test message was that when commenting a few minutes ago I got a big red error message.
This will be worth the wait!
As a fellow Canadian I’m sorry we cannot look forward to snow footage, but I understand what you explained about the batteries. (But what if you added battery heater blankets?)
Hi Tom – what did the “big red error message” say?
As for snow, I’ll probably give it a shot in the winter for a few minutes here and there, here in Montreal we’ve had some very odd winters, and sometimes we have had green grass in January. I just won’t be able to keep it outside too long.
Battery blankets would probably work, but of course, they would reduce the battery life as they consume power too.
super stoked on ardupilot, I can almost see my DIY way point drone with retractable claw hand to bring me beers while I lounge in the yard, in the sun, oh i’ll be the neighbourhood envy
ps your stuff has really helped guide me to some interesting areas, thank you!
I’d appreciate an approximate reference price of the cost of all the items needed.
Nevertheless, thanks again for all your communication and for sharing your know-how.
This sounds like a great project. I’ve been saving the chassis and wheels of an R/C Monster Truck just for THIS REASON! I was into Robotic a few years ago and have a 2 wheel self balancer and walking 17DOF humanoid-bot, but never had the programming knowledge to improve on them…so now it’ll happen as I absorb all the info from your Rover project.
I have been thinking about an outdoor robot for several years. I live on a gravel road about 1/2 mile from my mail box. I would like to make a bot that can go along the road and collect my mail. I have thought about various methods of unloading the box, but I think the simplest solution is to just exchange the box, leaving an empty box and bringing home the full one. The issues are navigation, the road curves a bit and goes over a couple hills. I think GPS is probably useful, but at the box location I… Read more »
ServoCity has a 6 wheel independent suspension rover chassis, available also from Robot Shop. 6 motors, one for each 5″ wheel. The base is 18″ long and with 5″ wheels, could be a lawnmower! Expensive though – $600+ CAD before tax. It’s a beauty though. If I do this project, it will be with this part or something close.
Hi Bill. Good to see you are still working on your projects. Your tutorials on arduino have been very helpful
GPS-RTK may be a challenge because it involves the guts of the GPS data interpretation. I am listing some references that may be useful to the project based on
efforts and experiments by others:
Inside GNSS Media & Research September/October 2021 Page 40 (mydigitalpublication.com)
RTK GPS Driven Mower : 16 Steps – Instructables
and last but not least:
Sub Centimeter GPS With RTKLIB : 4 Steps – Instructables (Probably the most useful to the project)
Hi Bill, do you have an idea when you have time to continue this great Project?