Amity Institute Of Training & Development

IOT in Insurance Sector: Today and in the Future

About: Mr. Peter Pugh-Jones leads a team of expert industry Consultants across Europe, Middle East, Africa and Asia Pacific Region for the Global IoT division at SAS which is the leader in Business Analytics Software and Services and the largest Independent vendor in the Business Intelligence market. During a Career spanning of over four decades, Peter has worked all over the World in diverse roles and Technology sectors and Software industries. A passionate believer of utilizing the right tools and Technology to deliver the best value and outcomes, Peter considers himself as a life-long learner with an optimistic view of future of Technology. With the convergence of Robotics, Analytics and Artificial Intelligence (AI), IoT is on the cusp for the fourth Industrial Revolution. He believes there has never been a more exciting or important time to be working in the advanced Internet space as we begin the journey to operationalize Artificial Intelligence and transform all lives.

Click below to watch the recorded conversation.

 AITD: Mr. Peter Pugh-Jones is based out of the United Kingdom. In fact, in this session, Peter will be talking about IoT in the Insurance industry and how advances in AI and IoT are driving Innovation for the sector and forcing insurers to consider new ways to define Policies and the Revenue models. Peter will also share examples of his experience working with customers in different areas of the Insurance industry such as Construction, Shipping, Transportation and Automotive. If I talk about Internet of Things, Internet of Things will radically change the World in coming years and Network devices will be substantial part of this change.

In 2010, people owned 12.5 billion Network devices. It is estimated that we are nearing or having already reached 50 million mark. IoT has entered everyday lives of Consumers around the Globe and transformed Business models across Industries. The Environment brings opportunities for Insurance as well to develop new products, open new distribution channels and extend their role to include Prediction, Prevention and Assistance. In fact by 2025, it is expected that the Insurance industry has the potential to automate 25% of its Processes, especially manual processes like Claim Processing, Underwriting, Customer Service and Policy Administration using AI and Machine Learning. Now, IoT Technology with the proliferation of data from sensors and smart devices present opportunities for Insurance to improve Underwriting and enhance personalization of products and services. The integration of IoT can also help in Loss Mitigation and prevention through Behavioural Modification and active alerts. According to Forbes, IoT could help Insurance cut the cost of claims processed by 30% and also lower premiums for Consumer accordingly.

Now, before we start we would like to know more about you and your Organisation? What has been the vision of SAS in attaining Leadership position in IoT Analytic, solutions and what has been the importance and significance of IoT in the insurance sector?

Mr. Peter Pugh-Jones: To help me tell just a little bit more information about SAS and the areas we focus on, I thought it might be nice to share just a few slides to help discuss the subject a bit more.

So, I am going to share a couple of slides, just while we do this part and then I will turn it off again in a moment.

So, to start with one of the Philosophies of SAS, so we are as you mentioned earlier, one of the largest Independent Software Company, currently by Revenue size in the World which is I think quite an interesting achievement with around about three billion dollars Revenue on a yearly basis. But, to get to that point, how we got there is focusing purely on an area of expertise within Software, which of course, is Analytics and what we do is we focus on helping people do what they need to do deriving value from Analytics in the life cycle of data so as the data is gathered as the data flows through their systems doing the best we can to optimize where Analytics can help the most and so to do that I just wanted to mention a couple of things on this slide.

We start by saying, Analytics is our core. We created something called an Analytics Life Cycle and that allows us to help people think about looking at the data figuring out how they work with that data and then building Analytics models and other things they might want intelligence from it that they can take and persist somewhere and they might persist that Analytics inside a platform. So, one of the things SAS creates is an Analytics platform, which people can pick up and embed in their other platforms, such as an IoT platform, so you might have a platform of services around IoT you might take the SAS Analytics platform and embed it in there so that you can expand the value you get from the data that you are working with and then beyond that you might also take advantage of some of our own solutions that we have created especially in the Manufacturing, Banking and Insurance sectors where people utilize solutions we have built from our knowledge of 45 plus years of working in the Analytics space to make life easier for them to utilize and gain Intelligence from the Analytics that they create.

We have solutions for things like IFRS 9 we have solutions for Risk Management and obviously those things are used all around the World at leading Companies in those spaces. If we go and have a look at a little bit more about SAS, I am sure many people on the call have heard of the Company before, since we have been around a while, but may not be aware of the fact that we actually have Offices in most major Cities around the World. we have a development R&D sites in the US which is obviously our Head Office, but we also have a R&D facility in India. We have some in the North of England and now in Scotland and we also have R&D facilities in a few other parts of the World as well. We are about 15000 people strong around the Globe and some 6000 of those people are PhD Mathematicians or Statisticians, so that allows us to be able to help our Customers figure out what they want to do as best as possible with their Analytics.

So, with all those things in mind, I am then going to just do a little Introduction around the importance of data, because for us, as an Analytics Company it is really all about data.

And, I like to use this picture which I am not going to name but to show that I am sure some people have seen that show before and I am going to use the word ‘Evolution’ to help explain why I have got that picture in there, so we are talking about Evolution of Technology which leads to Evolution of Growth in data.

So, this slide is going to start by positioning something that some people on the call may have never seen before. It is actually the very first portable or known as the first portable Computer that was called an Osborne-1 and it was originally launched in 1981. You can see that portable is perhaps a relative term. I mean that that's quite a chunky piece of equipment and it is worth mentioning this discussion of course is about connected devices and I would say that the nearest thing to being connected that had was if you use the keyboard with your hands then you know it's kind of connected.

But the reason, I have got it there is because I want to compare it to something that came into the marketplace only 27 years later and to do that we will pop it up on the screen so those of you on the call I am sure recognize an iPhone, the first iPhone in 2007 and the reason I am showing you the two is because of that expansive Growth. So that first Computer is actually would you believe at launch of the iPhone that the Osborne-1 adjusted for Inflation was about ten times more expensive than the iPhone was, when it launched and obviously that's to do with Economies of scale to do with Production of scale and so on and the savings in the Technology industry as it grew.

But it's also worth mentioning that the iPhone is 500 times smaller and is a hundred times faster and it's actually also just out of interest, it weighs 100 times less so all very straightforward numbers to think about, okay, it's actually 500 times smaller. So with those things in mind we are seeing a huge change from the people that started with that first portable computer to the people that work with those iPhones. Now, why am I drawing attention to that? Well, there is one other thing to mention that the iPhone is essentially on unless you turn it off; it is always connected to something. It's always on the Wi-Fi or on the Internet or if you like on your cell on your cell services; so it's on all the time and one of the things I want to do is borrow a slide from one of our partners called Cisco,

who I am sure many people know that was published I think quite a while ago now, around 2015. This slide comes from and as you mentioned earlier, you can see the Growth of devices that are connected to the Internet over time and this is terrific! Now, it is worth mentioning that this is what we call connected devices and I will come back to that in a minute when we talk about IoT and connected devices because there is a slight difference we need to just make sure we talk about it.

But, I like this slide for lots of reasons partly because these are very funny devices on here like a connected ‘toothbrush’. I am not entirely sure how valuable that really is but the real point about this is that only half a billion Computers of any kind were connected to the Internet in the early 2000s but only a few years later you can see that in fact by 2016, we are talking many-many more eight and a half billion plus of devices and the kinds of devices changed suddenly from large Computers; you have got light bulbs, you have got fridges, you have got ovens. You have got so Wind Power, you have got traffic lights, you have got all sorts of interesting devices that may or may not be successful in the marketplace but the point is they generate data and even more recently, we have got, door locks I mean I have a door lock that I can open with Wi-Fi right for to get in and out of my house. I have to have the gadget obviously because I work in that space. So this is really important because all of these devices create a lot of data and it's a well-known Statistic that somewhere in about 2015 there was a study done that showed that some 127 devices are added to the Internet every second around the World!

AITD: In fact, you are talking about 50 billion in 2020 as per McKinsey report; it was supposed to reach in 2025.

So, in fact we have added more devices we are ahead of target?

Mr. Peter Pugh-Jones: Yes, absolutely its insane how quickly that's escalating now!

Why do we care about that? Because of the whole concept of Analytics, so, if you have lots of things creating lots of data and you don't have anything helping you look at that data, how are you ever going to get any value from that data? So, that's why Analytics is so vital; that's why we play such an important role in the value chain of that data from creation to value.

And on the flip side of it, it is also worth mentioning that if you don't have any data or the right kind of data then there is not really much you can do ‘analytically’. You don't need to have data to be able to come up with models that could then add value to your sources so there is a heavy careful balance you have to make between the two. So in the old days and before sort of 2010, the general rule of thumb was well let's just grab all the data which makes me laugh right now because in the old days, you could almost say well I will just get myself a Hadoop Distribution somewhere.

I will put it in my ‘Data Centre’ or maybe I will use this new thing in 2010. I will use this new thing called the Cloud and I will stick stuff in the Cloud and see what I can do, once I have collected it all. It's a bit like thinking about saying “oh, just download the Internet and I will keep a copy of it on my hard drive in case I need it for later, right”!

It's a silly concept but that was what people used to do. They used to say, “let's collect it all and then we will analyse it when we have got it all”. But, the World isn't going to work that way anymore because the average Internet user consumes about one and a half Gigabytes of data a day and if you are my daughter by the way, it's way more than that as I have discovered right because she watches lots of streaming Movies. What do you keep right? Single Hospital can produce three Terabytes of data a day on average; modern cars like a Tesla generates about four terabytes of data a day most of which is not used or even like it doesn't even leave the car at the moment it's not really doing very much. An airplane, you can imagine generates an awful lot of data, 40 terabytes and a large factory, a petabyte of data a day. So, that concept of “hey let's just save it all isn't really valid anymore, right”?

So, one of the ways you deal with this is that you come up with new disruptive ways to deal with that Technology so we have got I mean this is one of our examples, we have got a solution called SAS Analytics for IoT which handily also has the same letters in its name as our key concept which is Artificial Intelligence of things and we like to talk about this. Being the way that we help people derive greater value from the data that might be coming from all sorts of places whatever the situation you are in and so that's why I wanted to just make sure I was able to tell you a little bit about that so that people get an understanding of where we are coming from as a Company in terms of our IoT knowledge but also in terms of how that then applies to things in in the Insurance space.

AITD: A couple of things like the iPhone is 500 times smaller and 100 times faster about the data and I remember when we started, we used to have 10 GB hard disk and used to say plenty of hard disk.

Now, we have even one Terabyte and that even then it is less because of data we are storing, storing and we don't know what to do about it?

Mr. Peter Pugh-Jones: It's really funny you should say that because I went with some of our newer members of the SAS Company to an event about a year or two ago; there was a sort of pop-up Museum in London and it was a Museum of Computer Technologies and I am walking around that Museum with people in their 20s who recently just started work and they go “wow, look at that I have never seen one of those before and I turned to what it was a floppy disk and I said I have used those before”. It made me feel like a dinosaur!

AITD: So, tell me what are the latest IoT trends in Health, Life and general Insurance because we are talking about Insurance sector and Health is which is very important during this Pandemic?

Mr. Peter Pugh-Jones: There are so many areas of disruption and transformations going on, especially in the Health space, which of course affects Insurance. But, before I start sharing some information about that space, it's important just to address the concept of IoT in that space and stay connected. So, I have been talking mostly about what I call connected devices rather than specifically IoT, it fits under the IoT umbrella but IoT is and most people will know is generally defined as machines that talk to each other without human interaction so the general rule is, if it's a machine that interacts with humans or requires interaction with a human, then it's not considered generally IoT as a purest definition, but I like to consider human interaction devices as IoT devices as well and that's because for us as an Analytics Company, we need to think about all connected devices as sources of data and if you are wearing an Apple watch like mine, it is a wearable device; it has loads of sensors on it and yes it requires interaction from me to do certain things but other things happen automatically and could gather data automatically.

So, therefore to me, it's still something that I consider to be a sort of IoT related device and that also by the way accounts for vehicles because a car currently requires human input to drive it's still considered more of a connected device then it is a IoT device. If that makes sense, but on it there are thousands of sensors and those do things automatically so therefore it's still to me relevant. Now, the reason I want to draw that distinction is because from here on I am mostly talking about connected devices generally and that includes devices that have human interaction so when we talk about Healthcare, the sorts of things that we are and that we have been working on seeing in the marketplace are generally around how it is with the growing population in the World and in some parts of the World, where we have an ageing population and less younger people, how can we allow people to stay at home longer if they want to and work Independently feel Independent at home and maybe provide them with additional security and Insurance around without feeling like we are watching them right because that's one of the problems with Analytics and Automation is people feeling like “well hang on, that's encroaching on my personal space”.

So, one way you can do that in the Health space is with things like audio; so we do have examples, which I will show in a moment of utilizing audio signals to process things like people walking around a room and they are boiling the kettle have they done that at the same time every day, are they therefore going through their normal routines and if so then everything's fine. If they are not going through their normal routines, is there something that we can do to check that they are okay that sort of concept right and we can have a look at?

Now, the other kinds of things in Healthcare, of course, which are the most obvious ones and that SAS is very heavily involved in is helping the sheer weight of diagnosis of patients, in say, for example, scans or CAT scans or where people have done X-rays right and for an average set of X-rays of say lungs or livers could be a very reading. High pile of things for an expert to look through to detect problems with people's lungs or livers but if you can use Analytics or Artificial Intelligence to maybe run through those images first instead of having a huge pile of things for one Doctor to look through, you could maybe reduce that down to just the ones that the AI have detected as maybe needing further Research right.

So, that's a really good example of how in the Health space, you can use Analytics and Artificial Intelligence to assist and certainly when that relates to Insurance Policies of course then you are talking about helping perhaps predict or reduce claims by predicting problems sooner than perhaps you could otherwise do and therefore catching people's illnesses and maybe providing proactive care in advance right which is something that is really very much on people's radars.

I am obviously not the World's leading expert on Health itself but on the application of IoT in those Environments is we have done quite a lot of things. So, Healthcare is definitely a space like that and then of course you have got safety people in Construction; there's Insurance related to people's working habits and what they are doing so in the Construction space, you might use cameras, you might use geo-fencing to make sure that people are carefully avoiding areas of danger.

Maybe you are alerting them in some way, if they go into an area where a crane could be dropping equipment or something like this and there's quite a lot of people implementing capabilities like that which are again linked to Insurance Policies right if people are to adhere to those rules in the Environment, then the Insurance is valid. So, Healthcare definitely is an area of straight Growth and general Insurance I mean you probably know there are Insurance Companies right now that offer Life Insurance Policies and also Health related treatment Policies, private Health Policies based on how well you comply with fitness routines. For example, on an Apple watch, if you wear your watch and you do the certain amount of exercise per month; you don't have to pay as much for that Apple watch than you might otherwise have to pay. It's a really good example of getting gamification right and to reach out to people to reach targets and keep and actually maintain some level of Health and fitness, which is probably a really good thing in the modern World.

AITD: In fact, very interesting aspects you have brought about. One thing which really touched deeply to me was Health and Safety of elderly people and why this is important.

Because these days, we have seen that the parents are alone and the children are abroad or they are working in different Cities and they are continuously checking about their health and they are worried and they want to know, how they are caring every day in terms of their daily routine so nothing is amiss?

So, that way, it is so interesting, and so important I would say because as the population is ageing, these things will be very handy to understand.

Mr. Peter Pugh-Jones: Actually, in the last couple of years, it's been very hard, hasn't it, to visit people, in person, right? So, there's been a significant advancement of people experimenting with Technologies like that in the last year or two, having realized that it really is an urgent thing to start, to resolve right because if we have another situation like COVID again, we really want to make sure we are prepared better for not being able to see people in person as much. That's a real impetus; it's providing a real impetus in the Healthcare and in the Insurance space to try and find ways of helping that and in the care play space and of helping that be easier should it happen again.

AITD: And moreover, like it is all integrated, all the devices are there so even for anything you need to integrate that information which is coming through. It is so handy to get this information which helps the other person to disseminate information as well as to understand what is going on. So, it's a very interesting that indication part I think is the best.

So, what development of IoT services for Ecosystem has a long time investment future capabilities and these are some of things which you have said so what I am asking is what is your opinion on Insurance should see the development IoT services for developing this Ecosystem in a long-term fashion?

Mr. Peter Pugh-Jones: So, some of the major Insurers that I am aware of and certainly in the UK, the ones that I am familiar with are definitely experimenting. As we talked about, we are able to link Technology to personal Policies, those kinds of things.

So, that's very good, but I think there is a new sort of disruptive element coming in the Insurance World. It's been around a little while but it's referred to as a InsureTech and it is definitely a space where there are start-ups being created that will come in and help disrupt the sort of the way that Insurance is done today and the way that it's been done for a long time. It's worth mentioning the principles of an Insurance Policy were created in the early 1700s, right through 1706 or something like that was the first actual Insurance Policy defined right for Life Insurance and I think if we think about that that's based on the principles and understanding of a world in 1706, only 50 years before did they realize that the Earth, the Universe didn't revolve around the Earth, so with that sort of thing in mind maybe it's good to re-examine and disrupt and think about the Insurance space today.

And, I think there's some really obvious stories out there which are very commonly discussed in the modern World. I can give you an example. If you take a Tesla vehicle and once you pick it up, you buy a Tesla you pick it up from the from the factory and you drive it home and you think “gosh, that's fun, I wonder what I can do to tweak and improve my Tesla”, so you have got your Insurance Policy you have taken out for the vehicle that rolled off the factory line but maybe you think “oh, I wonder what what's this mode here, oh I can turn it into what do they call it, I think it's absurd mode or something like that” and it makes the performance of the engine and the vehicle, not the engine, the motor in the vehicle and everything about the vehicle changed entirely from the original Insurance Policy and you can do it by pressing a couple of buttons on your phone.

So, how does the Insurance World react to that in in this modern World that we are in? How does the insurance world think about autonomous vehicles in in 10, 15 years’ time, is that going to be a standard? How do they think about managing the Insurance and whose risk and faults and responsibility is different, an instance happening, so I think, that kind of disruption is coming anyway because of the Technical advancements and I think Insurance Companies obviously have to move with the times to figure it out. It's going to be a fascinating time and I think InsureTech is definitely going to bring disruption to that space but in a good way.

AITD: What you are talking about is a very interesting part that is InsureTech, which we are seeing a lot of disruption in the Banking also the fin-tech Companies coming in. One thing which is very critical which is coming out of all this is that how can IoT impact market Growth with a rising popularity of usage based Insurance solutions in the Insurance sector.

As you said the first Policy was issued in 1706 and the parameters of risk assessment and all were very different what they are today. So, usage-based insurance is something which is a very interesting thing and I would like to know more about it?

Mr. Peter Pugh-Jones: I have only really been exposed to the usage of it and also how we can apply IoT capabilities or Analytics of the IoT involved in it. So, one of the things we could do is bring in some of our Insurance specialists to have a deeper discussion in the future around their thinking in that space but, for sure, I mean what I see about using space Insurance over the last couple of years has been fascinating because my daughter has just passed her driving test and she had to take out Insurance for her vehicle and I think the Insurance that she was quoted for a per year was about 2000 pounds a year, which is obviously excessively high from my perspective, I think, “gosh, it's that's a lot of money, right”?

But, if we agreed to have a black box fitted in her vehicle by the Insurance Company, they would cut that down by half right so that allows her to pay only a thousand, still a lot of money, but anyway it's a thousand pounds a year and let's face it, I am paying for that right?

But, anyway the point being, that they are monitoring over time using Analytics and Statistical Analysis, the performance of her driving they are not going to just cancel her Policy, if she exceeds the speed limit on any given day; they are going to look at characteristics over time and then over a three month period or a month period or whatever it is that they want to do; they are going to take into account all the metrics about her driving, maybe compare her to other drivers of the same age and then come up with a Policy over the year that will be suitable for her style of driving next year and I can see a time where my daughter will experience Insurance Policies that might change based on your usage without you having to go onto a website and change the data.

If you move house or you decide that you don't keep the car in the garage anymore or maybe you park it on the road at the moment but then you buy a new house, you park it on the driveway, you are required at the moment to go in somewhere and change all of that. Ring them up and talk about it all but actually, what they can do is that automatically really if they wanted to right now with Telematics coming off vehicles and GPS and location stuff and what great Customer experience, would that be if they say, “oh, we have noticed you now keep your car in the driveway so we can adjust your Policy to this amount if you agree let us know”, that's customer service right and that's moving towards the customer satisfaction side of things that people might like right. So, of course there's another side to that which could be the side that people don't like.

AITD: Just one example which came to my mind is as I was talking to somebody in the Insurance sector and it is said that, nowadays, there are no Intermediaries. You are not logging in, you are doing digital, you are searching for the policies and you are buying on your own, and the device which you are logging in also makes a lot of difference suppose you are logging in from iPhone that will decide how much premium how much Insurance you can be given.

Mr. Peter Pugh-Jones: It's fascinating! So, many more metrics that they take into account you are absolutely right, I mean yes, how are you even connecting to their website is something that they will consider right as part of the Policy assessment right? So, yes you are absolutely right, it's a fascinating area and I mean we provide service in the Insurance industry, we helped you credit scoring on claims and also on taking out Policies and it's a fascinating area to work in and for sure I hope it makes the experience from for people better because of it but there's so much more they could do they could automate a lot more of that I think based on data available, either from the vehicles themselves or from people's actual devices as long as the people are happy with that being used for those purposes right.

AITD: You were mentioning about your daughter getting a driving license which is something to be cherished as it's something we need to have a party once you get this License. So we come to a very important which you mentioned briefly is Auto Telematics.

I was just going through some articles where they were talking about how many times you brake, the driver applies brake, what is the tire pressure all these things, they have sensors which actually talk about deciding about the premium and know what kind of Insurance needs to be given to.

So, how does Auto Telematics allow Insurers to provide value-added services like driver feedback, theft prevention and road assistance?

Mr. Peter Pugh-Jones: In some areas of Insurance, we talked about earlier on an average car in the modern era generating about I think I said 40 Terabytes of data a day; a lot of that data doesn't really leave the vehicle and is not yet used for much beyond anything except historical review by an Engineer when you take it in for service right and there's definitely Insurance Companies are still at the beginning of their usage of that data and still figuring out the right way to make use of it for helping with Policy definitions.

So, some of it, they do quite well, other areas they are still negotiating. For example, there are two kinds of Telematics; there's the pure Telematics coming off the vehicle itself so that's where you have got the Manufacturer putting inside in the vehicle the Computer equipment that's in there already that generates lots and lots of data. Then you have got the sort of the hybrid that most Insurance Companies use today which is like my daughter's scenario where in fact they have designed with a partner some sort of block black box that they are going to put in the vehicle somewhere in the vehicle and it just collects a few things independently. It doesn't really read stuff off the vehicle itself it's sort of collecting things independently such as GPS speed momentum and a few other things, they could probably detect things like braking and stuff I am not entirely sure everything that's on those devices but it doesn't necessarily read the data directly off the vehicle so I think there's some there's still some work to be done in some of those areas right because obviously there's much richer data available in the telematics that comes from the actual vehicle itself rather than these new devices that are separate and added in separately if that makes sense.

But, most Insurance Companies tend to use those separate devices at the moment but there are Companies out there now that are partnering with the Manufacturers and the vehicle Manufacturers to provide methodologies for anyone interested, viz. Insurance Companies to gather that Telematics off the vehicles directly without needing a specialized box that has been designed for just specific things and, I think, that is a major advancement, that will become a very useful way for Insurers to be able to make more value from the data.

And, it's going to be awhile. I am not sure if any of the Insurers are actually reading directly from Telematics data off the vehicle. Yet, most of them still fit their own box and that's partly because of I think Rules and Regulation and Governance, they know what's in that box they fitted it themselves. It passed certain Rules and tests; so therefore, they can do it whereas of course if they are connecting to the raw data on the car they need someone else to really tidy that data up first and they need to be sure that that's being tied up in a way that meets all those requirements. Making sure people's personal data is protected and so on.

But, once you have those things you can move into a world which is really fascinating and we have been working in in that space for some time there are plenty of Customers we have worked with and partners we have worked with who always often ask us really interesting questions so we had one where they wanted to know about vehicles passing a particular stadium in the United States Football Stadium during particular times of day because of the congestion and the transportation problems and that was used a combination of geo-fences around the Stadium itself and then data reading off certain manufactured vehicles where they were sharing the information and that was a really interesting study that allowed us to help them optimize closures of road management of the things around the Stadium during sporting events.

Then, you have got the other side of that where you are maybe utilising the data or Customer Intelligence reasons and perhaps what you want to do is, as long as and this is where it becomes a little bit tricky you don't want people to feel like that you are being intrusive, right? I will give you an example, right, if I am driving my car past a Starbucks, right and my phone pops up with anything I shouldn't look at my phone right that's the first thing but if for some reason I am able to see the screen of my phone what I don't really know if I would like very much is my Insurance Company saying, “hey we're partnered with Starbucks, here's a free voucher for you to go into Starbucks and buy a coffee”. I might find that a bit, “it's like, oh hang on a minute you are really spying on me here”.

I don't know if I would like that but on the other side if I get a pop-up that says, “hey, we have noticed that you are feeling a bit drowsy why don't you pop into Starbucks and grab a coffee on us because we can see that you are feeling a bit sleepy from your driving habit maybe I would appreciate that, right”?

So, it just depends on how you do it right and I think that's the subtlety that Artificial Intelligence can definitely help with rights to make sure they are using the right methodology to communicate with you for the right reasons and make you feel like it's something that providing you benefit right and so that's another example and then and then finally I will show some slides just to give you some additional info.

So, this is more to do with asset performance and management but let me just share this and I will go to the slide for it.

So, this is SAS working with, can you see I mean these are just some web pages I scraped just to give people a little bit of background information but this is about maximizing the vehicle up time for fleet of Volvo and MACK is the brand in the United States vehicles. As they move around the US, doing their thing and one of the challenges they have is maximizing or improving the service time and ensuring that they do the service intervals at a time that's most suitable right and one of the challenges they gave us was well okay we have got a Telematics Gateway; this is a Volvo themselves, they have got a Telematics Gateway, they receive this data from on-board devices either from the actual vehicle itself or from ODB devices that they have plugged in and what they wanted to do was find a way to use Analytics to help improve the forwarding of information to what they call case creation support.

So, this means they are seeing errors coming off the vehicle they want to know how important that error is? Is it something that we need to send the vehicle in for immediate service or replacement or is it something that we can maybe wait until the next scheduled service for the vehicle? So, that we can optimize when it needs to happen and this has a continuous loop so if there is something really important flagged; it will get past pass through the this flow which is really a flow in the middle there of we call it a directed graph streaming assisted in memory flow where the data transforms through various different things gets checked through different filters and rules and also Analytics models and then outputs value that says up forward down to the case creation support team that there's a ticket needing to be created for this to be serviced and send an email to the customer that they need to come in for a service or they need to have this part replaced before because it's fairly critical.

So, it's a very straight-forward. And, this is a very high level by the way. But, one of the things that's really important is that it can continuously learn over time as the as the messages and the data coming off those vehicles changes.

I just want to take you to the next one. Because, just to give you an idea of the sort of scales we're talking about here and why we use streaming Analytics becomes clear because we want to make sure that it takes the shortest time as possible to get a ticket created if there's one needed based on the current scenarios is coming of that vehicle.

So, there is a sort of Customer Service Level Agreement; there is an SLA of 60 seconds to ensure that when there is a fault occurring within 60 seconds. It's been passed through the system and it's decided whether there's something that needs to be done and at this stage in the Project, we are working with about eight Terabytes of data, by now there is 71 billion plus different measures being looked at all the time through these tools through these streaming environments and you can see that over time they have detected about a billion faults or more why is that because at the moment that's actually supporting a fleet of vehicles across the US of about 350,000 and more and it gets bigger all the time obviously as they bring more and more of this online.

Why have they done this? Well, one of the problems is that obviously you get errors all the time off equipment on vehicles and so on and one of the things they were doing was receiving information that something needed to be fixed. When, in fact, maybe it didn't really need to be fixed so we have helped them using something like this reduce the false positives they were receiving by something like 44% and we have also helped them improve the diagnostic time by 70%.

So, that means, much quicker diagnostics, much faster diagnostics and therefore reducing the time to repair something but also reducing the frequency at which something is not available to transport product around the US. So, it's a really important to use cases and stories about using Telematics in a way that improves the ability of your assets to do the job that they need to do and obviously there is a ‘knock-on’ effect to that for Insurance, because you are talking about Insurance of parts, you are talking about Insurance of the vehicles, you are talking about if you have to replace parts, there are Manufacturers’ Warranties, there are other kinds of Insurance involved and all of those things can be impacted positively as a result of implementing something like this. So, hopefully that's a useful overview of an example of Telematics being used in a very positive way to add value with Analytics.

AITD: I am amazed to see by what percentage the repair time is reduced, by what percentage the other things; diagnostic time is reduced by 70% which is amazing. So, many things and it is evolving as you rightly said we are evolving these are some of the examples which are giving us so much of benefits so much of interesting data which we can collate which we can integrate to reduce so many things reduce human pain reduce time and efforts everything obviously it is very important to know all these things.

Mr. Peter Pugh-Jones: I mentioned, that I had some examples of audio being processed so just so that people can get a sense of what that means, I thought, I would just show you.

So, these are a couple of the pictures of audio signals and these are audio from heartbeats and it's just to give you some examples. So, there's a difference here between a regular and a regular heartbeat and what you can do is pass the actual audio track the sound through a streaming engine like the one we were just talking about with Volvo which will basically look for different patterns in the data and this is a visualization if you like of those different patterns and if there's an irregular heartbeat or something wrong with it in this example here they are using something called short time four year transform to help them do that detection if you can imagine that happens in real time so you could subscribe a streaming Analytics engine to the audio of a monitor detecting heartbeats maybe it's somebody in their home maybe it's someone in a medical scenario where you don't want to have to you can maybe you don't have enough medical staff to continuously monitor them and this could be useful to help give you early detection or early knowledge about something not going correctly with that person's heart.

I can, if you want to hear it but it might be a bit annoying. So that's a wheezing cough and that's a picture of a wheezing cough this one is a dry cough and this one is a wet cough you can see there's I am sorry about all the cough noises. Let's come off that slide because it runs on a bit. I will stop it from running too long because it can get a bit annoying. I find but the point is these are pictures of those kinds of coughing and again if you use this as a way to help understand what you can do with live cough processing of that kind of signal you can very quickly help diagnose things and of course these are being used now this sort of capability is being used in test scenarios in Japan, where they have got those first level GP Doctors which are basically Robots and they sit there in front of you, you do something like cough and they can give you a first assessment and then they can send that assessment to a real Medical person who can then go ah we see this is correct, this is a wheezing cough that means that this could be a way to treat it.

So, simple things like that but fascinating from a Healthcare perspective, but also of course fascinating perhaps from an Insurance perspective because it allows as long as it's used correctly to help people improve their Health and help people manage their Health better, then I think, Insurance is a valid place to think about how they could utilize that.

AITD: The most critical aspect and Health is the diagnostics part and if something is there to diagnose like you talk about heartbeat whether it's normal, it's lower, so that it gives an indication and the same thing with the cough, the sound which is coming out whether it's a dry cuff or it's a different type of cough which we have and it gives a clear indication and it helps in diagnosing for Doctors; it happens because I had issue with my son a couple of months ago we were struggling and for a month he was struggling and we were not able to diagnose but actually the problem is.

So, as you rightly said it is diagnostics and this is where IoT will be very helpful for a common person in in the Health sector, which is so important even today in Pandemic. We have seen that people are suffering so much and this this will be a very big boon rather Health sector. It has been such an interesting discussion with you and I wish that we can go on and on but the possibility of time and we will call it a series and this is one of them and we will call you again for different sectors, rather this is the Insurance sector. Then, we will discuss about more of other sector may be Manufacturing, maybe Telecom where all IoT will be interesting.

Before we end, just wanted to ask you that what the future of IoT is? If you may tell us more about it how it is advancing and how it is evolving?

Mr. Peter Pugh-Jones: A very quick answer to that within the Insurance industry, specifically, is what you can see we focus on. We focus on helping IoT be more useful by adding Analytics to it but putting the Analytics in the right place to be of the most value. So, in other words, if you build models you can then put them somewhere maybe it's near to the device that's registering maybe it's on the device maybe it's on the watch maybe it's in the phone if it wherever it needs to be to add the most value and I think in the Insurance space that's the sort of thing that is going to change quite a lot but there's a lot of adaption to happen first and they do I think for Insurance.

They are going to be and they are ready but they are going to need to think about the disruption that this kinds of Technology bring to the very way that they do things right. It's no longer a case that you can assess someone's Insurance risk one day and then a year later just reassess them you are going to need in the future to continually reassess people based on changing data and over time and but I think also people need to be accepting that that's the way that Insurance will work for them in the future.

Overall, outside of insurance as a broader subject, the future of IoT is to become a sort of everyday accepted thing in our lives wearable Technology. Even embedded Technology, I mean there are companies working on contact lenses that can adjust over time without you having to have operations there are people working on contact lenses that have video screens in them. It's going to just expand massively and the impact on Society I hope will be as an optimist I hope will be a fantastic result in time if used correctly.

AITD: Absolutely I am sure with so much so many devices being connected and such data being integrated I am sure there will be hundreds of applications which will be evolving for a period of time which will be so helpful for the mankind and people as such and will help in all the sectors whether it's Insurance, Construction, Telecom and Auto there are so many sectors, Health, where IoT will be a most important thing.

Click below to watch the recorded conversation.

 

Click here to know 'Corporate Learning Solutions'

Credits:

Moderated by Ashish Sahu

 

Explore our solutions - Click here