HyperCat is a new open specification developed by a consortium of 40 UK-based companies, educational establishments and local authorities that's designed to spur on the development of the Internet of Things (IoT).
Backed by £6.4 million of Technology Strategy Board funding, it works by creating an online catalog tagged with metadata that can be read by other IoT devices. This allows for interoperability, meaning streetlights could theoretically automatically communicate with parking sensors instead of only other streetlights (for example).
To find out more, TechRadar Pro spoke to Justin Anderson, CEO and co-founder at IoT vendor Flexeye, one of the companies involved in HyperCat's development.
The standard
TechRadar Pro: What is HyperCat?
Justin Anderson: HyperCat is a specification for the Internet of Things that, given the appropriate push by government, will become a standard.
It will allow anybody who is using some kind of sensor to sense something – whether it's their environment, home, temperature, use of location or speed of travel – and then make that data available in a way that others could pick up and make use of that information.
Clearly, there's a lot of information that you want to be careful about publishing. You might have information about yourself that you're happy to publish, but you only need to put out on HyperCat the bit that you're happy for everybody to publish.
The possibilities
TRP: Can you provide an example of a scenario that HyperCat would make possible?
JA: If you're driving your car, you might be happy for the location of that car to be accessible to others – or you might not. You might not want anybody to know where you're driving, but you might be happy for it to be provided if, by releasing that information, your insurance policy comes down.
So by giving some information to your insurance company, you could reduce your premium. Or by providing some information to the AA, they'll come and find you faster. Or by providing some information that links your car to your driving license, you can get your tax disc automatically sent to you.
Being able to let your things talk to other people's systems has lots of potential advantages and can make our lives much easier. Lots of systems now are starting to make data more and more available. The challenge is that when you make that data available, you need to make sure that anybody who wants to come and read it understands how to read it.
We need a common language to make sense of the data, so HyperCat lays down some basic principles by which systems can understand, read and identify the information associated with a 'thing'.
Balancing scales
TRP: Why is it important that machines and systems know what Internet-connected 'things' are?
JA: Let me back off for a moment to look more broadly at the development of technology solutions. Most recently we've seen a lot of technology coming out which is being driven by some very large vendors that run things from end-to-end, which could be considered a vertical silo.
That could be everything from the sensor to the application and everything across their operating system, including their applications, which are fully locked down. Companies like AlertMe, which have systems that work within your house, work out how to piece all these bits together to create a full solution. That particular one was licensed to British Gas as part of their Hive offering that manages the heating in your house.
They hit a problem as they scaled which is how to manage the entire system from end-to-end - they wanted to pull in best of breed bits of technology to do that for them - which took them toward a requirement called modularisation, which is about breaking it up into different pieces.
For modularisation to work, you have to have the different pieces talking to each other in a language that's easily readable and understandable. So HyperCat is a way of being able to define the information that's coming from a thing within a catalog, which is the Cat part of it, to let people know what information they're reading.
The consortium
TRP: What parties were involved in its creation?
JA: There were 50 companies that were involved, both large and small. They included ARM and BT, and small companies and startups that you might not have heard of. There was also half a dozen different universities. We also had local authorities in Westminster City Council and Guildford Borough Council. We had also had the Government - public and private sector - and technology companies like IBM. Flexeye, as a UK SME, was one the key players at leading the standard and taking it forward.
But the key was that they all had to agree on something, and trying to get 50 different companies to agree on something is much more difficult than it sounds because ultimately, if you use one company's system, then everybody will be happy and it will work well. But everybody puts forward their own system, so that doesn't work and you have to agree on a way of doing it.
Connected society
TRP: Who will HyperCat benefit?
JA: One of those people is the everyday citizen who is living in a world that will one day see tens of billions of sensors on this planet. Over the next decade, as those sensors are connected, the world's order will change.
There is something fundamentally critical that's happening to how our connected society will unfold. It could be that as that happens, there are some very large companies that will connect all these devices together and provide you with a way of controlling those.
There are some efficiencies in doing that, and they have a lot of resources that they could apply to doing that, but we could also look forward and imagine the world controlled by a small number of very large companies that could control everything that's going on. George Orwell springs to mind - that's not the way to foster innovation.
There's another way it could work, which is that these 50 billion sensors are able to be controlled and monitored by lots of different technologies that collectively combined to allow us as citizens to control, manage and monitor our environment to allow our governments to have a level of governance over our environment without putting control into the hands of a small number of commercially-driven organisations that have the power to control the world.
As an SME, I'm for the little guy, and I'm a believer that we have an opportunity to modularise IoT and allow different players to do their best bits within this system, and HyperCat allows that.
It allow the small guys to compete with somebody with $100 billion on their balance sheet and come up with solutions that tie together information for my car and house and deliver it to me as a citizen without me being tied to paying a higher cost to a large international company that controls what I do.
UK advantage
TRP: How will HyperCat drive innovation and potentially boost the UK's tech potential against overseas rivals?
JA: First of all, it's not in the America's interest to drive standards when they've got such an embedded base of very large corporations that are very powerful in themselves and also very powerful at lobbying governments.
On the other hand, the Brits have a very highly innovative set of citizens who have always done a great job of innovating as individuals or as small companies, and that's what has fed our economy.
We're very well positioned to be able to tie together an awful lot of information that collectively will allow us to compete with an organisation that has $100 billion dollars on its balance sheet. And it's only by creating this collaborative, innovative environment that we can genuinely compete against other nations, but at the moment we're very well positioned to do that.
We can take this next step and lead the world. What does that mean? If we can drive more than our fair share of value-add over this next period and achieve that $100 billion, that would be a great boost for British industry, jobs, taxes and UK exports.
Privacy issues
TRP: Are there any privacy and security issues related to everybody's data being accessible by every other system?
JA: It's an important point. Who should decide how people's data is used? Fundamentally, if it's data that's associated with me as an individual, I should be in control of how that data is used.
There are times where I also believe that a society requires its government to be able to have access to certain information to be able to protect its citizens, which is one of the first roles of government. It needs to be able to use information, but getting that balance right is really important.
Let's look at the corporation: at what point should it have responsibility for owning, managing and controlling data about me? Clearly there are places where that should be allowed, but we need to make sure that they're the appropriate places.
If it's easier for me to get hold of an insurance policy, I might be happy for a number of insurance companies to get information that they might use in certain circumstances to be able to make that process easier for me. I might give it to them myself over the phone, but being able to automatically provide it to them has some benefit.
This ownership of data cross the public sector, private sector and individual is critically important.
Communication issue
TRP: How important is the manner that corporations communicate that cost-reward factor to consumers?
JA: I think what will need to happen at every step that data is transferred from one player, stakeholder or organisation to another, that there needs to be a policy associated with that data as it's transferred.
We've seen some problems recently with systems put in place to support the NHS, which has seen different vendors responsible for different parts of the system that haven't worked out how to pass the data across and swap policies from one provider to another, which is where it falls down.
There are other aspects of security that lie around making sure nobody has tampered with the sensor, then there is the physical security that sits on a chip, as well as the security of making sure there are no viruses on your database. There are all different kinds of security, but making sure you have the right policy associated with the use of that data and the authorisation of that data for use within different applications is very important.
Vendor lock-in?
TRP: If vendors adopt the HyperCat standard, would that reduce the chance of vendor 'lock-in' that we've experienced in the cloud world?
JA: If the principles by which the information is communicated between one party and another are open standards and we all agree that this is the way we'll communicate those policies, then that takes the lock-in out.
There will be vendors that come along with solutions that will manage that whole process better than others, but so long as they're able to plugged in or out dependant on how good a job they're doing, rather than plugged in or out because they effectively control the whole stack, that will be the key.
Can you replace IBM with FlexEye, or visa versa, if you need to? That's the question. if you can't then we have problems, which leads to organisations working to their own strengths. If FlexEye has a better solution for managing the authorisation process than IBM, then it should be the one that's managing that process for somebody, but you should be able to switch them out if necessary.
Training machines
TRP: How is HyperCat, and by extension IoT going to improve our everyday lives?
JA: Without any question we're moving into a world where a lot of tasks will be performed that humans might have had to perform in the past. The other side of the question is what about jobs? Are there going to be any jobs left for us to do?
It's important to recognise that these machines are very good at processing information once they've been provided with the context, but they need to be trained, and we need to continue to train them. Our brains have 100 billion neurone inside them that can process lots of different things and intuitively can get to answers that machines can't get to.
There's always going to be a requirement to move to ask the human at a certain point in time, so we're going to be come increasingly responsible for managing the exceptions that come along, and they will happen all the time and our brains will be needed to do that.
For consumers, one of the key things is that if our products that we buy are going to have sensors in them, it will change the way we might work with our products and the way that we interact with stuff around us, but it'll also change the way that we interface with those companies that sell things to us.
If you look at the industrial space, at Roles Royce, they sell their aircraft engines not as engines, but based on how many hours they're in the air. Their whole business model is making sure they can service the aircraft staying in the air.
When we as consumers move to a business model that isn't 'we're going to buy a thing', but 'we're going to buy a thing that I will pay for in some way shape or another based on usage or something else', then that changes the relationship I have and ensures that the things I buy will be good quality, because if they're not the vendors aren't going to make any money, so it'll drive toward a high quality use of technology.
We'll also find that the products given to us are more compliant with a broad range of policies that are important for us to be complying with as citizens. We clearly need to be looking at a more energy efficient planet, and IoT gives us a way to drive higher levels of energy efficiency. We can switch off lights automatically, or manage streetlights, or manage congestion in cities by helping us find parking spots.
TRP: When will we see HyperCat-compatible devices?
JA: They're already out there. The important thing is that it's not so much the device, it's how that device makes itself readable.
Playing nicely
TRP: How will HyperCat play with other IoT standards, such as Neul's Weightless standard for "white space" communications?
JA: We're focusing on the piece that makes the information on the data and thing available as it goes from analogue to digital, whereas Neul and Weightless are focused on spectrum. There are a whole load of challenges around spectrum as we move from licensed to unlicensed.
There are operators who want to control that spectrum as much as they can - they don't particularly like unlicensed spectrum being used and there's a big drive to try and bring that under control in some way - but at the moment it's not.
You've got a whole bunch of spectrum that's unlicensed that allows us to innovate in a way that reduces that cost, and long may that last.
http://ift.tt/1pGuzpR
No comments:
Post a Comment