Last updated: 23 March 2015
At Mobile World Congress earlier this month, we showed off a working prototype of an automated 3G Bushfire Alert System, the winning idea from our recent #IoTMaker Challenge.
The idea came from Jason Mitcheson, who works for Australian farm technology provider Observant. We spoke to Jason to find out more about what inspired the idea for the Bushfire Alert System, his thoughts about the future of IoT and M2M technology and where his idea might go next…
What inspired you to come up with the idea?
I was on a short camping holiday a few months ago, which here in Australia is bushfire season. Camping in bushfire season means checking fire bans, and driving past a lot of highway signs to do with bushfires. My mind started to wander as I rode in the back seat of the car reading all the signs, and I thought about the possibility of using Gemalto Java modules as compact fire detectors for remote regions.
How did it feel to see your idea become an actual prototype?
It was extremely humbling to see the working prototype. I felt very inspired and grateful that this all happened. Seeing an idle thought that you had in the back seat of a car come to life on the other side of the world is quite a trip!
The proof of concept was built on a Cinterion Concept Board powered by J2ME. You’re a Java developer too – what makes Java well suited for the IoT?
A couple of things:
– The ability to handle problems without crashing of a whole system. IoT systems are usually remote, and become a huge pain to debug or repair if there is something unexpected which happens. In Java you can catch an exception and have other parts of your system continue as normal.
– The Java language has concurrency support. This is an obvious major benefit over typical native languages if you want concurrent tasks. It’s easier to use Java threads than move to an RTOS in C, for example.
What’s the most exciting piece of IoT technology that you’ve seen or heard of?
Connection of vehicles to the internet. Especially drones; right now all consumer drone technology uses radio or local networks. I’m also working on an internet enabled flight controller using the Cinterion EHS6 Java module, which I call “Highstream”.
How do you work? What does your dev workspace look like?
I used to have 3 screens when I had a PC; I’ve calmed down a bit now. I use a Macbook and 1 other monitor, plus a docked iPad. I emulate Windows 10 using VirtualBox to run the Concept Board development environment. I write a lot of JavaScript so I’ll usually have WebStorm open and/or IntelliJ for writing Java code. My other project involves JavaScript, a Raspberry Pi, and the Cinterion module so the development environment gets very busy!
You work for a company which makes technology solutions for the farming industry. Do you see IoT technology having an impact here too? What are some examples of how M2M/IoT technology can benefit the farming industry?
Definitely. We are already at the stage where every piece of equipment on a large farm can be connected to the internet. It is having and will have a massive impact on farm labour costs, energy saving, and effective crop water management.
The architecture of IoT on a farm might be a little different than most people imagine, though. The actual equipment (pumps, soil moisture sensors, energy meters etc.) does not come manufactured to be able to connect to the internet; rather they must be plugged into a specialized piece of hardware which talks to the equipment using a serial connection, and then forwards messages to/from the hardware to the cloud at set intervals – rather than the equipment connecting directly to the internet and constantly being active.
– Live stock (cows, sheep etc) need to drink water, and can suffer undue stress if the level of the water tank on the farm runs dry. IoT technology can monitor these levels around the clock to ensure it never goes too low.
– Crops use a lot of water. Soil moisture probes let us estimate how much water is needed to keep the crop healthy, and at what time in the future it will be needed by. M2M/IoT technology means this prediction and analysis can scale to hundreds of probes because it can be managed by cloud software.
– Electric fences can be monitored automatically to detect breaks or faults. This improves the integrity of the fence, thereby improving safety for both the live stock and the nearby roads.
– Larger farms have many paddocks that need to be watered. Manual watering means additional labour costs for the farm as the paddocks can be quite far away from the homestead and a long drive is required each time. Connecting the irrigation infrastructure to an IoT cloud means that the watering can be carried out by one person, from one computer at any location.
Smart cities get a lot of media coverage but do you think there’s just as much potential for the connected countryside?
Definitely – if not more so. The countryside is all about production: everything is active. Electric fences, vehicles, pumps, solar panels, live stock; all of these things are constantly active so the value from monitoring them has a huge impact. The same can’t quite be said of smart cities as living in a city is more about passive consumption.
What advice would you give anyone thinking about embarking on their own IoT projects?
Start small, and try and base as much as you can on kits like the Cinterion Concept Board, and/or the Raspberry Pi where you have a community you can look to for existing solutions and to ask for help. When you build the cloud side of your IoT project, I cannot stress enough that you will need to build in debugging and network introspection tools right from the beginning. If you don’t have a good handle on every single byte/message being sent around your system to/from your remote units then you will get stuck quite easily!
What do you think the future holds for M2M/IoT innovation?
I think the participation of a wider developer demographic will be a big part of the future. Java and backend-JavaScript based IoT technology will bring easier development, a bigger community, and quicker prototyping to a space which is traditionally a bit slower and harder to develop.
Can you imagine a KickStarter campaign to get your Bushfire alert off the ground!?
Definitely! Watch this space…
Many thanks to Jason for his time speaking to us, and congratulations again to him on winning the IoTMaker Challenge. If you have any thoughts on IoT applications for the countryside, want to share how your Java dev workspace works for you, or have another idea for an IoT project, let us know in the comments.
