How AI helped a startup to achieve ultimate simplicity
Yes, you heard it right. Simplicity it was and it is the aim of mankind, since the very first beginning. It is such an important part of our learning and understanding process, that even Albert Einstein expressed it with a famous quote :
If you can’t explain it to a six year old, you don’t understand it yourself.
In the following paragraphs, I am going to share with you, how we were targeting ‘simplicity’ from the very first days of product, Flare, and which amount of ‘sophistication’ it requires to get there.
What is Flare ?
Flare is the name of our home security device, that it’s meant to provide you the safety, confidence and comfort to ignore it, while you enjoy your life.
It is packed with sensors, while latest software technologies powers its circuit boards, in order to ensure no human error factor affects the device on its mission of offering top safety to our users.
For all the geeky lovers out there, we have revealed more informations about the hardware components on our blog — feel free to unwrap them.
How did we got started?
If you’re going to ask an iPhone user what simplicity means to him, he will answer by showing you his phone. Usage, design, performance — it’s pretty straightforward, right?
If you try to apply the same principle to home security, you will find out that there’s still a lot to be gained, especially when it comes to affordability and the nerves of the end-users! Entering codes several times a day, staying alert for notifications, managing false alarms and turning yourself into a freaked-out person: That’s exactly what we don’t wish for our customers.
Imagine a real-life scenario of waking up in the middle of the night or getting disturbed during a conference, just because your home security system has triggered a false alarm. Or how about receiving an alarming notification, just because the wind is blowing in the curtains, or your cat crouches in front of the cam?
Being a witness of such scenarios, and seeing an opportunity in the middle of the chaos, it simply lead us to the conclusion that
no one should spend time on their security system.
That’s how we started the BuddyGuard project, with a courageous idea — believing in a self-managing system, that is able to detect and analyse burglaries before they happen and take counter-measures on it’s own, without needing user intervention.
It may seem like a lot, but since we are living in really interesting technological times, we believe putting together all those functions in one device is possible, and that’s what we are doing.
What’s the price for simplicity?
In order to accomplish this ultimate goal, we had to solve 3 main problems:
automatic activation/deactivation of the system
recognising the threats with maximum accuracy
taking counter-measures on it’s own.
We first started to look for a solution by investigating expensive hardware radars, that could potentially tell the distance to a person and the profile of that person. In general they are providing good results on ‘detecting’ persons, but, unfortunately, it only solves only one side of the problem — we still need to ‘authenticate, recognise’ the owners, while setting them apart from intruders.
At this moment, we’ve started to redirect our attention towards software, for algorithms that could potentially compensate the hardware limitations.
Artificial intelligence is still one of the hot topics on the researchers tables, no matter if it’s algorithms are used for hand-written recognition, face detection or any other interesting tasks.
We started to integrate AI for our home security scenarios and soon we got a full fledged set of algorithms for face recognition, voice analysis, schedule learning and pets detection. For the sake of accuracy we’ve added phone location geofencing as well, so that device knows automatically when you arrived in the home area.
These algorithms are providing the device with a deep understanding of the outside context, such as who did arrived at home, what is the cause of the noise or motion in the apartment, when it has to automatically activate itself or who is the intruder in your house.Having all those insights at hand, it helps to take better decisions for each case.
Besides the benefits for the user of having a self-sufficient device, there is a great benefit on our side as well, since we are no longer subject to the ‘human error’ factor — forgetting to activate/deactivate the system in the traditional way.
Offering simplicity is hard. The sophistication of the algorithms behind, combined with the hardware components, makes everything a difficult to build, yet beautiful system. But we would rather go the extra mile, than settle down for something less innovative.
What’s next ?
We are preparing a Kickstarter campaign for Flare, with only 5 days to go until the official launch day, on May 12!
We are looking forward of making the change towards more ‘simple, yet powerful systems’, together with readers like you!
Thus, we are offering 25% discount when we launch our product, for early subscribers like you. Be the first to pledge on Kickstarter or buy the product later at a discount rate!
You can expect many more updates next week, including two beautiful feature videos, more photographs of Flare in a beautiful apartment in Berlin and a sneak peek of our official Kickstarter video!