Creating Systems of Intelligence

The world is rapidly changing. Many technologies are organically adopted, while the transitioning to some technologies is the need of the hour. With the rapid change in technologies, business models too are fast changing to adapt to these changes. Companies wanting to sell a standalone product is now passe; they now aspire to sell products as a service. This revolutionary change in business model is a reality in many sectors today. Technology can also help companies improve the productivity of their people, equipment and existing technologies. To demonstrate the new business models and the way technology is impacting the bottom line, Vishnu Bhavaraju, Manufacturing Industry Director, Microsoft India, spoke at length at a CII event in a session titled Business Excellence - Futuristic Technologies. In Part two of our coverage of the session, Corporate Citizen brings you the excerpts from his riveting speech

I want to start off by saying that my colleagues in the US were colleagues of my fellow panellist, Rohit Gupta, Head-Technology Center for Analytics & Software Engineering at Thyssenkrupp India, in one of their biggest businesses, which is the elevators business. They figured out a way of selling these elevators using a very different model. They now sell elevators as an up-time as a service. They don’t sell elevators to buildings anymore. Now they install elevators in buildings where you don’t pay anything upfront but you pay as a consumption model, just like you pay for your newspaper at the end of every month. What they did is use Internet of things (IoT) and Artificial intelligence (AI) and technologies like that to collect data and can now figure out the possibilities of what can be done with that data. Initially, the idea was to get the data to understand how well the elevators were performing in customer locations. Once they did that, they formulated that why sell elevators as a product when you can sell them as a service. It’s a very innovative model and it’s a very successful model as well. From elevators, you can bring this down to the simplest of things. I was in Delhi recently with one of my customers who is a tyre manufacturer and they are looking at launching tyre resistance. They are thinking of no longer selling tyres as a standalone product. They will look at tires now as Cost Per Kilometer (CPKM) The customer will pay them x paisa per kilometre, and take the tyre without making any payment upfront.

"Earlier when we were adopting ERP, we were creating a system of recording. We were doing a transaction. Now what we are doing is creating systems of intelligence"

What is happening is that people and companies are looking at creating a competitive advantage. How do I serve my customer better than my competitor, is their primary question. What enables these possibilities is digital technologies. My question to everyone is “When did we start adopting ERP solutions?” A couple of decades ago perhaps. When did we start adopting computers in India? A few more decades ago. Were computers and ERP systems analogue or digital? They were digital. So, digital was always there. What is happening today that makes people so excited about digital now? What has changed? If we were doing digital systems 20 years ago, if we were implementing digital ERP and CRP systems 20 years ago, what is new about digital today? This is how I would put it. Earlier when we were adopting ERP, what we were creating is a system of recording. We were doing a transaction. We were manufacturing something and selling it to someone. To manufacture, we needed a production holder, which we were earlier recording in books and later in ERP or computer system. Then we were selling it to the customer, so we recorded the customer details in books and later in systems. What we were doing was creating a system of records. Now what we are doing is creating systems of intelligence. What ThyssenKrupp did was to create a system of intelligence. They collected the kind of data which wasn’t possible to be collected earlier, and they are using that data to create a system of intelligence which can identify when a part of an elevator would need to be repaired. So they have the confidence to guarantee the up-time of an elevator. So data is creating systems of intelligence. That is the main differentiator between what we were doing decades earlier and what we are doing now. Now coming from the most advanced ‘product as a service’ perspective, let me give you a very simple example from India. A coal trader approached us. They said we import coals of different grades, which we can call Grade A and Grade B. Grade A being of a higher quality and Grade B being slightly lower in quality. They import coal and they sell it to coal consumers, whether they are blast furnaces or factories, etc. The business model was to import coal at a lesser price and then sell it in India for a profit, just like any other trade. The challenge they had was that there were leakages. In any other manufacturing, there is always something called stock reconciliation, which is how much you actually sold. This helps you understand how much of a product you bought, how much you sold and how much should be left at the end of it. You can always collate the data to find the exact numbers. Now imagine the coal trader, where it is not feasible to count the exact quantities as it would be in packaged products. Coal requires a lot of space. Trucks come; they dump the coal and leave. Then another set of trucks come, they pick the coal and they take it away. In such a situation, how do you reconcile the stock? There is no way to do stock reconciliation. Which means, there is a lot of chance of leakage. This could be intentional or accidental. Say I buy five truckloads of Grade B coal. I will send five trucks to the trader, and it has to be loaded. If by mistake instead of loading Grade B coal, the person loads Grade A coal, or I strike some ‘arrangement’ with the person so he intentionally loads Grade A coal even though I paid for Grade B, it is a huge loss to the coal trader. Look at the equipment that they have today. The truck comes; there is an excavator which loads the coal. So how do you prevent leakage using technology? We use RFID and other simple technologies. When the truck comes, it is matched to the excavator which will load coal into it. The truck is also matched to the Grade of coal it is there to collect. So through this real-time data collection, we can know which truck is coming from which customer, what grade of coal he has ordered, and which is the excavator which has been activated to load the coal into the truck. The weight of the empty truck is then compared to the weight of the loaded truck to capture the amount of coal loaded into it. Even then, there could be some error in weighing as the weighing scales may not be exactly accurate. So we installed a CCTV camera and we will use AI to convert the data into actual numbers. Using simple technology, so much can be done.

"What we see, are areas where we can improve productivity. The productivity of your machines and equipment and productivity of your technology"

Taking this to the next level-one of our customers is L&T. They have cranes and excavators and movers all over the country. They are doing construction projects everywhere. They have projects going on in Pune. They often lease cranes for work. They asked if they could track these assets. They started doing that and today they not only have visibility of where those assets are lying vacant but they can also see what the utilisation is to the extent that they can even see the weight that the crane is lifting. The crane manufacturer said that if you touch my electronics, I will void the warranty. So now what? So we put a tyre based monitoring system. The tyre would now tell us what was the difference in its pressure corresponding to the weight the crane was carrying. The point is that digital is about systems of intelligence. It could be about any business.

What we see, are areas where we can improve productivity. The productivity of your machines and equipment and productivity of your technology. When we look at the shop floor, where can digital technology help me with increasing productivity? We work with a lot of customers. People leave the information on the shop floor. This is usually very basic information. We get to know what jobs are running on a particular machine. If a machine goes down, which jobs will be impacted? Which customer orders take a hit? These are simple things which are required but it takes a lot of time to figure out these things. So, if you have heard of bots, they are very prevalent around the world today. Our customers have also figured out that this is the easiest way. If you deploy bots, either in the kiosk or in plants, people can question these bots and they can ask simple things. They can ask things like are spare parts available, how is the equipment they are using maintained in other plants? Not just that, people can use collaboration technologies. A company came to us which launched certain products. They said they wanted to reduce the launch time of the products. Now, who is involved in the launch of a product? Research and Development is involved as they are the ones who formulate the product, the supply chain system is involved, as they are the ones who can say whether there is enough supply of the product to match demand, the manufacturing shop floor is involved as they have to clarify whether they have enough space on the shop floor lines to manufacture the product, and the marketing team is involved as they have to decide what the market is for the product.

"If you see a scenario of quality testing using AI, if you want to adopt a new technology you can just take it on cloud and run it for a month to see if it works. If it doesn’t work, you can discard it"

But how do you put all these people together in one single place to collaborate? That’s where people collaboration technologies come in. What we see areas where we can improve productivity. The productivity of your machines and equipment and productivity of your technology. The third area is in the area of quality. We are working with manufacturing companies across sectors from solar panel manufacturers to glass manufacturers. They ask how they should reduce defects. You can use the cameras which are already there on the shop floor, identify defects and flag those defects. But there are challenges as well. For example, this glass manufacturing company came back and said ‘when we use the cameras to take pictures, there is a lot of reflection in the pictures due to the materials being glass. So how do we identify defects with this obstacle in the way?’ So now there is a human inspection which is being conducted. When the person who looks at the cameras to identify defects gets tired or fatigued, a lot of defects can creep in. But we can use the cameras to detect fatigue in workers on the shop floor.

The second thing I talked about is improving the productivity of machines. I gave the examples of L&T and coal traders and showed that even on the shop floor we can improve productivity. The third thing is that we are using technology to improve our productivity, and we must also increase the productivity of our technology itself. What I mean by this is that when you talk to manufacturing companies today, every company says that they are specialised in manufacturing certain products, whether they are cars, glass bottles, or other products. And they ask ‘How do we ensure that we don’t lose too much time and cost maintaining our systems?’ That’s where the cloud technologies come into play. Whatever the solutions, whether ERP or any other ones, they are being migrated to the cloud by companies. Just like other companies offer products as a service, cloud providers sell up-time as a service so that your application is always on. You don’t have to worry about buying a bigger machine to install these systems. Another way to make technology more productive is that when you look at adopting any of the digital technologies, instead of worrying about the cost of buying the new technology, worrying about installing it, and finding people to deploy it, the cloud technologies make everything available at a fraction of the cost. If you see a scenario of quality testing using AI, if you want to adopt a new technology, see if it works on your shop floor, you can just take it on cloud and run it for a month to see if it works. If it doesn’t work, you can discard it and focus on something else. That’s how technology can be productive.

These are the three areas I wanted to talk about, the productivity of people, equipment and technologies. This has been the focus as we see and talk to our manufacturing customers.