IoT – what’s working, what’s not, what’s still missing and what to look out for
There seems to be a deep divide and an uncrossable line between Industrial and consumer Internet of Things (IoT). This is the first thing that became apparent at our discussion on IoT, at the TCS DEX Event last evening. Industrial IoT addresses a set of mature needs and mature problems, with arguably better and cheaper technology, and enables the much-valued integration of IT and OT (Information Technology and Operational Technologies). Consumer IoT, on the other hand, is a brave new world where largely new needs are being identified and solved. By new needs, we mean issues and challenges that were not seen as problems that could be solved by technology, or perhaps even weren’t explicitly seen as problems. The fact that you couldn’t turn on your home heating as you were leaving office wasn’t a problem, and that your toothbrush couldn’t calculate your brushing efficiency didn’t keep you up at night either. It would be fair to say that the consumer IoT world, therefore, is a world of new opportunities to make life better, in ways we couldn’t imagine before.
Let’s look at the Industrial IoT first, though. There’s a reason why a lot of experts, including those like Shirish in our discussion yesterday referred to it as old wine in new bottles. Industrial IoT typically involves getting about the operating conditions of large machines, plants or supply chain elements, in order to bring costs down and improve visibility and predictability. Now, most industrial companies have been doing this or trying to, for ages. Most large plants already have sensors, and the data is being captured for similar purposes. SCADA (Supervisory Control & Data Acquisition) systems have typically been the way this has been done. But even in the industrial internet, the IoT model is bringing about a few changes that are significant.
First, the sheer range of sensors and sensor-enabled devices are exploding. It’s not just the very large and expensive equipment anymore but sensors are going into heating and cooling systems, moving parts, components and all kinds of environments. You can, of course, read the reports about IoT and sensor explosion (insert appropriate number of zeroes here). This is, in turn, exacerbating an existing problem. The challenge of industrial IoT has historically not been the lack of sensors, but the inability to capture and use the data already being generated by the machines. My colleague Rishabh, who is our resident expert on industrial IoT, suggests that only 0.5 % of all the operational data generated is actually used or stored and the rest was historically discarded. Now, with the IoT tools and big data capabilities, the remaining 99.5% is actually being captured, analyzed and used. This includes smart visualizations via VR headsets or AR glasses. A third key change is that the I in IoT stands for the Internet – which implies the use of public networks and a wider reach over low-cost networks. This means that at nominal cost, this data can be shared with an ecosystem of partners – from insurance companies, to maintenance teams, to customers. Think of a large shipping business like Maersk. Their IoT initiative now allows customers to track their containers across the world. This ability to push the data beyond traditional corporate networks, and utilise the Internet is a huge enabler of new services and value. Which leads to the fourth key difference – the emergence of new business models. The poster child for this is, of course, GE which uses the Predix platform to track aircraft engines and now leases rather than sells the engines, and controls the performance of the engines through the IoT platforms.
A critical aspect of the IT/OT integration and any business model change is that the ability to understand and manage the unique context of each business. Traditional IT has had the luxury of being the same or broadly similar in its challenges irrespective of the industry. Not so with operational technology and data. This is why implementing IoT solutions is far beyond the typical IT approach, and requires deep immersion into the operational issues and challenges – for example the specific heat and humidity conditions required for pharmaceutical supply chains, to the stress, temperature and conditions that need to be withstood by an aircraft engine, as well as the hours of operation, the cost of downtime, the availability of spare parts and the contractual obligations in place.
Through all of these, there are some common themes – making businesses processes simpler, and easier, making business decisions better and more predictable and also enabling the cultural change required to drive new operating, process or business models. After all the engineer who has spent a lifetime looking after a machine and can tell by the sound it makes that it is underperforming, does not want to be told that a system of sensors and data can do a better job than him in predicting and preventing breakdowns of the machine. This cultural resistance to new technology is common across almost every new technology and in every domain, so I’m not going to belabour the point here. The broader discussion and one that’s worth having is about the user experience being delivered by the IoT solutions, even in the industrial scenarios. Let’s come back to that later.
The other half of the IoT story is the consumer IoT narrative, and this is dominated by stories and anecdotes around everything from smart toothbrushes, to weighing scales, to refrigerators that can order your milk when it runs out. Or how about (as Cecile says) cocktail umbrellas that open and shut based on where in the world it’s raining. As with any consumer product, the magic lies not just in the functionality but in the emotional connects that the product can make with its users. Or between users. One of my favourite examples of this is the Goodnight Lamp created by Alex Deschamps-Sonsino. Consumer IoT products are all about finding worthwhile ways in which we can get consumers to simplify their lives, and find value through smarter and connected objects, in ways in which they couldn’t earlier.
Here too, there are great examples of business model changes – which include everything from City Bikes which can be left anywhere, to home appliances which can be tracked for their performance, by the manufacturer. Once again though its the context that drives the true value of IoT solutions. This is why players such as telecom businesses in some parts of the world have looked to play a much more active role in providing IoT based services, such as smart agriculture, in Japan, which goes as far as the ‘internet of cows’. The idea of ‘contextual integration’ is a very powerful one – where the same product could behave differently depending on the context. For instance, when customers were fleeing Hurricane Irma in Florida earlier in 2017, Tesla increased the range of its cars by upgrading the battery pack.
Given these great use cases, the question that begs an answer is, why aren’t IoT solutions more universal and what’s holding them back? Gartner suggested earlier this year at their Symposium that IoT pilots are many, but scaled solutions are few, and they called out security, complexity and cost escalations as some of the reasons for the lack of enough scaled solutions. Our anecdotal evidence suggests that the answers may lie in how the use cases are selected and how the project is set up and pitched to the leadership team or the board. Common issues seem to include overpromising and choosing overly ambitious use cases. Finding a simple opportunity with a clear and achievable business case is apparently the gold dust here. It seems quite intuitive but perhaps this is a case of ‘uncommon common sense’. A great example of finding value and a new business model in the consumer space, as quoted in the TCS 2015 global study of IoT “Internet of Things: The Complete Reimaginative Force” is HP, who created a “Printer as a service” and “Ink as a service” models by switching customers into a subscription service which prices by number of pages printed, and involved HP’s ability to track the ink usage and levels and replenish printer ink at the right time.
Having said all of this, there are still some key areas where as an industry IoT still needs more thinking. The first of these is the User Experience of IoT. This represents a major evolution for the discipline of user experience as this marks a significant move away from screens as well. It’s one thing to say we will have smart things – such as smart buildings and smart teacups, but how do we actually interact with a smart object. There hasn’t yet been enough exploration and clarity into this next phase of user experience. In the book, Designing Connected Products, Claire Rowland, one of the authors points out that this is ‘not a cookbook’ as it’s too early. No doubt that moment will come. Another provocative way of posing the same question is, does IoT need its Apple moment? A company which creates a next-generation product that changes the face of IoT and truly defines the standard that everybody else must live up to, in the IoT space? Many consumer electronics leaders display great products at CES – and one of them may well be the company that redefines IoT for us, but its safe to say it hasn’t happened yet. My colleague Ravi puts it succinctly when he says ‘it boils down to the solution understanding what is it I want to do…’
The other key area that I keep thinking about every time I read a prediction about gazillions of sensors and connected devices, is, are we creating a new legacy? Who is going to look after the 20 billion devices Gartner says will be in place by 2020? How is their lifecycle being managed? What is the cost of replacing and upgrading 20 billion devices? Many of which are embedded in machines, castings and literally set in stone? What is the environmental impact of this?
Whatever it is, we are likely to see more and more IoT products, platforms and solutions in our lives, both at work and at home, and behind the scenes in every service we consume – from trains and planes, to hotels and offices, and even in the water we drink. To me, though the place where we will truly see the best results of IoT will not be in our connected homes but instead in our connected cars. The automobile is already a ready platform, packed with sensors and computers, hundreds of components, with an existing network and a great need for contextual awareness. More and more companies are looking to deliver better experiences in the automobile – from Tesla to Rolls Royce, and we can expect the connected car to lead the way in the silent IoT revolution that is taking place as we speak.
This post is a reflection of the discussion we had at our office yesterday, on the topic of IoT Revolution. Many thanks to Abraham Joseph, Mark Challinor, Ritesh Jain, Vikram Kanwar, Damien Stephens and my colleagues in the room for making yesterday’s discussion an excellent and educational one.
As with all my posts, all opinions here are my own – and not reflective of or necessarily shared by my employers.