Unmanned Robotic Mower

written by mr337 on 2016-01-16

For Christmas I got an Arduino electric kit as a gift. Soon I was working through the tutorials making good progress. It wasn't too long afterwards I was like like "What else could I build?".

For new things I usually like to test the waters and do a bunch of research before going further. With the Arduino it was quite the opposite. Soon I started to have all kinds of plans and projects. So if going, go big!


How I came to lawn mowing actually originated from the farm. I do not find brush-hogging any fun. Inhaling fumes all day, getting body beat to death by rough tractor, and if your unlucky, no cab on the tractor soaking up all those mid summer rays. Depending on how much you are brush-hogging you can expect from a full day to a week or more. Huge miserable time sink.

Initially when planning safety became a huge factor. The thought of a 4 ton tractor cruising at full throttle with no control was an uncomfortable feeling. Amplified by 100 when your trial farm is next to a major interstate. So deciding that was a bit too much I had to scale down, thus the mower was born.

After doing some thinking I have come to a set of expectations and goals for a robotic mower:

  • Cut +90% or more of the lawn
  • Gas powered blade, for simplicity, with electronic drivetrain
  • Location aware with centimeter precision for optimal travel
  • Enough battery to mow one acre of lawn
  • Must not be impeded by non square lawns or obstacles
  • Follow programmed path

This should provide a mower that can do a yard by itself requiring an operator to make close cuts such as flowerbeds and next to fences.

Nice to have:

  • On-board charging unit for electric drivetrain
  • Automated obstacle routing - only provide electric fence
  • Optimal cutting paths using cell decomposing
  • Quick survey/discovery of new lawn

Operator allowed actions:

  • Start and kill motor
  • Position mower within lawn
  • Maintain oil and fuel levels (no auto fueling, fireball anyone?)

When assessing these requirements it soon became clear this is a big project. A good chance I won't meet half of those expectations. At minimum I'll learn a thing or two.

The main reason for this endeavor is to learn about electronics. I have experience on the software side, but almost zero experience on the hardware/electronics. In an effort to share the knowledge I plan on doing a series of blog posts on what issues and, hopefully, solutions encountered.

So far this is just a lot of talk. To get things moving and commit, I've already ordered the hardware and waiting for them to come in. In the meantime I'll start drafting the different systems and how I think everything will come together.