Big Brother in the Bathroom: AMA with Intel’s VP of IoT
Startup Scaleup kicks off their AMAs with Philip Moynagh, VP of Intel’s IoT Group
At the start of December, we launched our first Ask Me Anything (AMA) session with Intel’s VP of IoT, Philip Moynagh.
With our focus on IoT and thanks to our partners at the Ryan Academy, we were delighted to have the Vice-President of Intel’s Internet of Things group, Philip Moynagh who views IoT as the merging of the physical world with the digital economy that is set to make the internet as we know it look like an opening act.
Philip set the scene by providing some examples of how IoT is solving real-life problems that range from optimising the collection of rubbish, preventing rhino poaching and even monitoring behaviour at urinals in airports (apparently this is more important than you think).
To create sustainable solutions with impact, you must know the problem inside out, understand the market, speak your customers’ language.
We had a full hour of Q&A with Philip. Here is a small selection of the questions posed by our start-ups and Philip’s answers to them.
Q1: We recently came across your IoT developer zone, can you tell us whether this is this giving you any insights into how the market is developing?
Medium and large companies are more open than we think. Any large company is stupid to think that they can engage IoT on their own. At Intel, everyone is acutely aware that there are far more smart people outside Intel than inside, there are 7.3 billion people in the world, and you have to figure out where the next new and exciting thing is coming from.
I go to a lot of maker events, in local universities here, and there is no better way to meet people. All medium and large companies if they are interested in being in and staying in the tech business need to have a clue and watch really closely multiple vectors of which this is an important one, watching the maker market, what is interesting, what the developers are making, it is developers that are the people that make the difference, bringing together different pieces from the software, hardware and business.
Q2: How seriously do you think privacy issues will limit IoT growth especially if there are serious loopholes discovered in consumer-facing tech?
This has a left and a right-hand edge to it. From the left, we know how to do security. There is not a single one of us that hasn’t done a credit card transactions with our numbers flying around the internet. We move billions of euro around the world and you’ll see https so you know it is a secure connection. This is a bit too flippant but look, we know how to do security don’t get too obsessed about it.
However on the right-hand edge, unfortunately there are way too many stories of loopholes arising from the fact that the most common password in the world is 123456 through to poorly architected or just not architected at all for security. There are lots of smart people out there who will spend their evenings learning how to hack and you are as strong as your weakest link. From an Intel perspectives we have spent 7.5 billion dollars on McAfee for exactly this reason and the attack vectors are fascinating, just to see the number and the inventiveness.
The architecture isn’t as good as it should be, and we’re designing too many different things. It’s like driving into a city and changing all the traffic lights and too many are poorly done but through consortia [the issues are beginning to be resolved].
Q3: Many IoT objects are using devices like Arduino, Rasberry Pi, etc., because the users don’t need knowledge in hardware development. Is Intel providing solutions with similar characteristics?
I’m a big fan of both of those and many more, the more people that are facilitated to develop solutions and the more tools there are out there, the better.
I see the IoT very like the PC, it’s a standard architecture, there are rules that you need to follow in order to build a thing called a personal computer and if you do not follow those rules it will fall over and it will not run the stuff that people want to run on it. Following those rules, it’s such a generic and broad architecture that I can buy a thing that has been built to PC standards for 200 euro and I can buy one for 20,000 euro, the architecture is the same, the purpose and the types of components are the same. We haven’t hit that level of clarity [yet] on the Internet of Things, there are way too many custom solutions designed specifically for that vineyard or that lawn, if you exquisitely deign your solution for that lawn then when you bump into an opportunity in the sports arena, a vineyard or whatever, you can’t adapt it and you are redoing your development costs.
So find building blocks, standardised building blocks that are as secure, or are as manageable or can facilitate as much scale as you need for the markets that you are targeting
The straight answer is that they are fantastic tools and we are big fans of them, our versions are Galileo and Edison and there are a bunch more coming. We donate a lot of the stuff we build to universities and events and they are relatively cheap to buy. It is useful to be on the same architecture if you want to move up to atoms and then cores but for now what we are interested in are the folks who have experimented on that scale: how do you become an easy on man for those folks to provide scalable, secure, manageable, analytics-capable solutions? How do you scale them in a manner that the thing can be re-purposed with multiple apps or that the apps can address multiple things? That’s not the way IoT planet-wide is currently constructed and one of the two reasons we are not scaling is the absence of those standards.
Q4: What solutions are available for wireless communication between two small devices over a short distance? WiFi is not a good option as it uses too much power.
From an Intel perspective we are agnostic, of course we would love if that everyone used one wireless or wired means of communication but of course that is not the reality of how it all operates. There are plenty of open architecture venues with light protocols out there to try.
If you want to send a large piece of your information then the protocol around it in WiFi is going to cost you thousands of bytes in burden. If you’re only sending a byte that is intolerable, so you need to find the right protocol that is the most cost-effective/burden-effective way of doing it. There are hundreds of bytes protocols and tens of bytes now in place like LWM2M that if you put into a search engine you would get.
Personally I have a bit of a bias towards a 6LoWPAN environment, we have customers who have solved the specific type of problem with Bluetooth LE, with 6LoWPAN and there are customers that do it with WiFi, it is ugly and you need to put a lot of resources in but at least it works.
When you get that really perfect 3m solution working on BTLE or a lightweight M2M protocol, the minute you have a customer that wants the solution on a 30m scale you then have a problem, so just think about your target market.
It certainly is an exciting time for IoT as the momentum builds with everyone from the Indian government to large enterprises getting behind solutions, but start-ups need to stay aware of the business challenges the ‘thing’ creates for business models in IoT compared to screen internet. Thanks Philip for sharing your experience and insights.