Our summer project with a team of students from UTC Reading is designed to give them experience of creating an IoT product prototype from a customer's idea. The purpose is for them to experience a highly structured and organised approach with technical discipline which is hard to get any other way.
But what's the product? What are they going to make?
A Smart Cat Flap! Sound silly? The more I thought about it the more I realised it makes sense. Many cat owners allow their cats outdoor freedom but call them in when they're going out or at night (the vast majority of cat disappearances occur at night).
This is how that usually pans out... You're ready to go out but where are the cats, are they in? Now you search the house while someone else is shouting out of the door and shaking biscuits or rustling food packets maniacally. Oh no! one of them is out, you think, and you go to find them. The other one is interested by this and tries to follow you out. Having foiled that plan you find the one you're looking for and start fielding him or her back into the house but you have to navigate the entrance where the one who tried to escape previously is hiding invisibly right next to the door. You are literally herding cats and there must be a better way.
Attaching a tracking device to the cat springs to mind, it's obvious, why hasn't it been done? There are a number of pitfalls which come to mind as soon as you start thinking about it...
The more expensive and bulky (for a cat) end of your system is attached to the cat's colar.
Cats lose collars, all the time, because they're designed to come undone if they're snagged.
You'll need a battery.
A device like this uses a relatively large amount of power and batteries go flat. We need a better solution.
We came up with a smart cat flap which knows whether your cat(s) is (are) in or out by detecting which way it last went through and can tell you in an instant just by glancing at the app, perhaps even just by looking at the apps icon.
Better still, we can add auxiliary sensors which could, for example, be mounted on a window used as an entry and exit point, or by their food bowl, the back door, litter tray, wherever they might get close to a sensor, this would increase reliability but also opens other possibilities (more on that). The only thing attached to the cat's collar and liable to loss is an RFID chip. These are tiny, really tiny, and incredibly cheap!
We wanted to be even smarter though. what if you could set up a schedule so that the cat flap is locked inwards, outwards, or both in line with your pre-set schedule? Want more? Set the schedule on a per cat basis so your big old tom can do as he pleases but the little skittish one is only able to get out during daylight, and when you're in, for example.
How about doing more with those axillary sensors? We'd be able to track and plot behaviours like knowing about 'denied' attempts at getting in or out. You might need to know that both cats are eating if one has been ill or see if one is chasing another off its food, know if one has been in the shed. Almost any place you want to know if a cat is going could have a sensor placed to pick up the cat's presence and log it.
We can set alerts, not just logging this data but allowing real-time information. If your cat has ever had a day or two long excursion, you'll know the worry. Set an alert to let you know immediately if that cat is back. It'll be locked in, of course, because you'll have set the flap to do so. You might want an instant alert for any fairly precise location.
The team have a lot of coding skill already and we want to offer them an insight into the disciplined approach we have to follow as professional product prototypers and introduce them to as many of our other skills as possible, commercial research and component selection, testing and de-risking, custom PCB design and dealing with the commercial supply of small batches, assembly, API driven annexing of systems and interfaces, the more we can squeeze in, the better!
We've defined the idea as above, just like a customer with an idea would to us. We've given them a technical lead (Steve) to steer them though the process and access to all of our engineers but the thinking will be theirs. They will be told how we do our research and given a few hints based on our experience, but they will do the research. We'll teach them a bit of the electronics required and introduce them to PCB design, we'll teach them the coding practices required to create clearly defined interfaces and to code different parts of the system simultaneously using those interfaces. We'll be introducing them to some cutting-edge cloud technologies and we'll teach them about 3D design and printing. Some of the work will need to be done in our lab so smaller sub groups will get to work in that environment as well.
Most of the team have some experience with coding (some have quite a lot), some experience with simple electronics and micro-controllers and they're incredibly smart and enthusiastic. This is going to be good!
Once again, thanks to UTC Reading for being the outstanding UTC they are and doing all the necessary background to make this happen for their students and Innovation Catalyst / Vital6 for providing work space, offering us flexibility of access to their facilities and taking on a considerable responsibility.