There has never been a better time to come up with creative ideas in ag that involve hardware and software, thanks to the mobile phone, maker movement, open source and enormous amount of free education online. Just about anyone with passion, grit and time (the hardest of all) can build just about anything.
I am not thanking the mobile phone because it gives us communication and access to apps etc, that’s all pretty boring and taken for granted now. I thank the mobile phone because it makes everything so damn cheap. Think mass production of GPS, cameras, wifi, gyros etc – economics work in our favor.
The maker movement is basically taking advantage of the situation of cheaper parts construction and packaging up parts especially for curious hobbyists. This makes sourcing parts easy for hands on prototyping, even in Australia. Also gives us access to cool technology like 3D printing.
Open source is generally something – such computer code, 3D model or information – that is licensed in a way that basically anyone can use it for free as long as they adhere to the license terms. Generally, large, successful open source projects have thousands in their ‘community’ and sometimes hundreds of contributing developers. What this means is a lot of the complex core work in many situations is already done. The APM autopilot code base is a fantastic example. Build a drone or rover OR EVEN A SUBMARINE and there is tested working autopilots ready to control them for free with great communities to help you out.
Now all this is really neat because people that have qualifications in say agronomy or experience in farm management that have real problems to solve can upskill and then have an overlapping understanding in agriculture and their associated technology of interest.
Once you have an understanding (or know where to look) of what is out there, it doesn’t take long before stacking together some cheap hardware with open source software ideas can become a reality.
Here are some ideas that may be more simple to build:
- Vehicle tracking devices
- Moisture probes
- Grain flow meters
- Weather stations
More ambitious ideas floating around in my head:
- RTK (repeatable, centimeter accurate GPS) is getting very cheap, and we have had high quality long range radios for a while (for sending data to and from base to rover). We could build a an open source autosteer controller for tractors as an alternative to options from Trimble and John Deere. We could use all open formats and generic hardware. See what Matt Reimer built using the APM autopilot controller. The end result is about as incredible as the support he got on the DIY Drones discussion forum. I expect RTK GPS prices to fall even further in the next couple years.
- Using our RTK GPS and rover version of the APM we could build small autonomous vehicles to do spot spraying. I took my first step a few years ago by building a small prototype. I have not got round to building a larger one (yet! – there is still time).
- We could mount WeedSeeker or WeedIT cameras on our open source autonomous vehicles or we could build our own! Computer vision and AI to ‘learn’ what objects look like is all the rage at the moment with some good open source software packages such as caffe. Rather than be limited to just whether an object has chlorophyll by a vegetation index, how great would it be to determine if it was grass or broadleaf. On board computers are fast and cheap and cameras are probably not as expensive as you think. This is getting very complex but not impossible.
- Lets take it one step further. Why not have a drone tethered to the rover. It could be up high constantly scanning for weeds and using some sort of smart travelling salesman algorithm find the most efficient way to spot spray all the weeds with the ground vehicle. Now this is getting somewhere albeit very ambitious.
All sound exciting? Where do you start?
- Remember – small steps and time.
- I learned to write computer code at Udacity for free by doing the Intro to Computer Science and Web Development. Both are a few years old now but still fantastic. edX now exists which looks exciting for high quality free online learning. I’d love to do a course in software machine integration, computer vision or AI next.
- Maybe hop straight to it and start working with Raspberry Pi or Arduino projects. My First project was a fixed wing UAV.
- Putting this post together I discovered Farm Hack – a website that basically shares this idea that it’s a great time to innovate in ag.
Do some reading, Google searches and have a think. You may just be able to make that thing your have been dreaming of yourself!