IoT: the shape of things to come

IoT Connected

There is little doubt that the Internet of Things is going to impact our lives in the future, for better or for worse, depending on which side of the fence you are sitting on. However, there are three major challenges we need to resolve before they can really start to transform our daily lives.

The first challenge is the network reach of IoT: most IoT devices (including sensors) need either a local WiFi network or operate using Bluetooth technologies. These bring a severe limitation on range and mean that a lot of IoT devices just miss the mark when it comes to their usefulness outside the home, office or factory. A solution up until now has been to use 3/4G networks, but that means fitting IoTs with batteries that require regular charging. This is fine for cars but not ideal for small sensors, GPS trackers, or wearables. In fact, using a mobile network for many IoTs is a bit of an over-kill: they are only designed to transmit and receive small amounts of information at various intervals. So, the long-term solution seems to me to build low-power, wide area mesh networks. LoRa is one such example that is vying for this space. KPN in the Netherlands is rolling out its own LoRa network over 2016 and 2017, which is likely to be subscription based. Some start-ups, like The Things Network, are trying to compete by setting up free LoRa networks in major cities. The reality is that some services will be for free whilst others, perhaps depending on heavy encryption and strong security, will warrant some form of subscription.

The second challenge is the result of solving the first and is not so easily resolved. How do we keep IoTs secure and ensure their identity and data are only accessible to authenticated machines and people? And how do we keep ourselves and our own IoT devices safe from the beady eyes and ears of others? This is a fragmentation issue that is becoming a wider challenge for all internet connected devices. It is about recognising devices, understanding what they’re doing and who they’re doing it to. Up until recently, most internet devices, such as your personal computer, laptop, tablet and mobile, have had a reasonably long and often predictable lifespan. They are also fairly straightforward to secure, especially when put behind a firewall. In the Internet of Things, each object could have a different lifespan. Some life-cycles will last for years, decades even, while others will last for just a few days or even minutes. Consider a trackable package; as soon as it arrives at its destination its identity will disappear from the Internet of Things. True Internet of Things will not be protected by traditional firewalls and secure point-to-point connections. Companies like ForgeRock are trying to address these issues but this is going to take some careful consideration to overcome privacy concerns dealing with personal data, particularly in healthcare.

The last challenge is how to get the most out of the Internet of Things: this is about automation. As we begin to incorporate more and more IoTs in to our daily lives we will not want to have to pay them constant attention; they will need to work for us. Our house lights should switch on and off automatically, and the lighting should be appropriate to our mood and what we are doing at the time. We also want our house to be at the right temperature when we’re home, and know that when we’re away that energy is not unnecessarily being wasted. We also want to know when our washing machine or dishwasher is no longer working optimally compared to the newest models on the market. How wonderful would it be if our regular shopping was ordered and delivered automatically for us, by our kitchen? Like most people, I also want to be a centenarian, so I don’t mind having my diet, health and fitness constantly monitored.

Nowadays, we tend to clutter our lives with technology that constantly disrupts our focus, constantly vies for our attention, and disengages us from the world around us. In my opinion, The Internet of Things will only truly start to come to life when we have built systems that can tap in to all the data that is being generated about us and our ‘things’ and can start to make sense of our daily routine so it can deliver value to us by automating some of the more mundane aspects of our lives. This is a luxury problem, but it’s one that could deliver enormous benefits to our environment, our health, and our mental well-being.

Ok, Computer: understanding me in the Internet of Things

Speech Recognition, Artificial Intelligence and mobile technologies have all been converging over the past decade to form a perfect storm which is now being underpinned by what we are calling the Internet of Things. Over the coming years we are going to be seeing new devices of all shapes and sizes. These will be in our homes, in our offices and factories, in our cars and most likely we’ll be wearing them; and we’ll definitely be carrying them! Most of all, these devices will not have keyboards as input, and will be operated by voice, touch or gesture.

This all sounds a bit like a scene out of Spielberg’s “Minority Report”, and there is no doubt there’s a long road ahead, but today, most people in the UK and The Netherlands own a Smartphone, and these are even more powerful than many laptops and computers 10 years ago, yet their screen sizes are barely large enough for us to do simple things like reading the news and responding to mail. Most people would not even contemplate doing their weekly shopping on them, yet for the first time, on average, we are spending more time looking at our mobile devices than our television. So it’s only natural to look at ways of reducing the friction of interacting with them.

At Capito Systems we are working at the convergence point of Speech Recognition and Artificial Intelligence and are learning how to apply these technologies to provide a more natural interface for people to use in their daily lives. To see how we’re positioned check out this recent article by Venture Beat and this infographic.

Nowadays, interacting with computing devices using voice is not just the holy grail for computer scientists. Technology has advanced far enough that it has real commercial value as well. There is a definite latent interest in consumers to use it, but I guess they’re weary of it because they need to see it work. First and foremost, the interaction needs to be natural for people to be comfortable doing it. For that, the device needs to be able to understand non-prescriptive language as well as be able to cope with accents and environment. It then needs to consistently respond with something useful that engages people to want to use it again. These are the core challenges facing our startup and I’m delighted to be giving a talk at ASAS 2015 on 14th October to demystify some of the complexity around this subject, as well as to explore the nuts and bolts of the platform we have been building to take away the heavy-lifting normally required to integrate natural language interfaces in to real-life applications.