The Hindu Business Line : Linking the shop floor to the top floor...

Chatting with Subu D Subramanian, Director and Senior Vice-President, Satyam Computer Services, is a very different experience for a journalist. Most industry captains expect you to fill in airtime, or conversation time, if you have asked for the appointment. And, rightly so too. But Subramanian came prepared as well. So, the `chat', ran beyond two hours, although we had anticipated only a 45-minute session. Excerpts:

What opportunities does the IT industry have in servicing manufacturing?

The India brand in IT is strong. So, it is the first choice of preference. We need to leverage that and then extend it. As a competitor, China can say, "I can do manufacturing and by the way I can do design also." Instead of following that, we can go the other way.

We can do the design and quality will be at its best because most of them will be automated beyond the design. So, when you say the design is the core, the mechanical and CAD/CAM processes have been largely automated. That is where we have superior skills from the automation stand point. You can support the whole design process.

The second thing is what enables collaborative engineering - such as product life cycle management and the like - where we could connect and distribute the design activities to multiple locations and then allow for integration, so that's again an area where we could articulate our capabilities.

The other area is digital manufacturing, which would enhance the productivity and improve the quality significantly. For, we can do the simulation of the manufacturing process upfront in an IT platform.

This is being extensively used in digital manufacturing in big names in manufacturing even before changing the set up in the CNC (computer numerical control) machine. Clearly, if we do it on the shop floor we need to do it on a trial-and-error basis. Then the downtime will be high. But, if you can simulate the whole thing and switch the line in no time because you tested and did the mock-up up front, the savings are high.

And, it's not the simulation of the product alone. You can even simulate the layout and the place to locate so we can improve the productivity with improved ergonomics and the like.

Lastly, manufacturing is itself going to be automated. That's the third area. First was design, and then the tools that enable the manufacturing, and now manufacturing itself.

The predictions are that companies globally are going to spend more on what they call the shop floor systems.

What it means is now, I can bring in more automation in the manufacturing process which will improve their vision, and then how I can leverage the manufacturing data into my planning purposes, so that's what integrating the shop floor to the top floor is, because today the ear piece is not connected to the shop floor systems.

The process control, the production control, the CMC machine, the data generated is not connected to the ERP and people don't know how to connect because only the shop floor people know how to operate the data generated and are only used for that purpose.

You have to give the output and input alone that goes in there?

Correct. For example, how many components it has processed and how much time it has taken, what is the status of the work in progress? That visibility is not directly linked to the ERP systems

So they manually key in the data so obviously there are time lag challenges.

There are levels to this. One is automation. The other is leveraging the data and then integrating that into the main supply chain so that would help them to predict and optimise the supply chain and have the real time view of the whole floor of the production itself

And, it's not just in one plant. Many companies have multiple plants globally. By interconnecting plant operations they will have a view of what component is not just lying in stores but lying in the progress and where it will be so which they can move quickly and then leverage and move it forward. So that's the kind of interconnection and collaboration and then integration with the ERP that's going to be the main stream which again is going to improve the efficiency of the overall manufacturing.

Do Indian vendors play a part in the automation of the manufacturing itself?

Yes, we are getting there: a few players, not all. In fact, we have recruited a few people who actually worked in a shop floor. I've recruited someone who was working in Ford and in Daimler Chrysler; that's the kind of team we have recruited globally and likewise we have people who have worked in Germany and Japan so we bring in the global perspective. Then there is a team in India, too. Spending is going to happen in this area.

When we project the value, the customer is able to appreciate this because the Accentures and IBMs of the world don't fit in this area. It clearly goes into the core computer science: how I can understand the process control and systems programming and how I can get back data from there and integrate it with the ERP? It is not like business systems it's a shop floor. It requires a high degree of IT services.

And these people are like domain experts?

We need to have both. We will have a domain expert to understand what the manufacturing process is and what systems and solutions they require. We will also require hardcore technical experts who can get into systems level programming.

So, when you say system level programming, rarely have we indulged in this in the past. It's always application or operating system, so is this the first of this kind?

Yes, but the catch is it is now being done in an isolated way. Process control, business applications and ERP deployment are all different applications.

From a client's viewpoint, these are three islands: it has a VP-engineering, VP-supply chain and then it has VP-production. The VP supply chain is always served by the CIO, but the CIO is no way related to the VP engineering and does not even get into the shop floor. Customers face the challenge of bringing all these together, from the engineering to the shop floor and then to the supply chain.

So you'll be able to participate in all the three, automation of the process, leveraging the data that comes out of it and connection to supply chain?

Yes, I'll tell you a classic case, which I myself experienced. I was into manufacturing before I moved to IT.

It starts typically with the product launch. The marketing folks talk of a new product, see scope for great market share, they do all the analysis, we need to launch and cut over for this new product. It then goes to engineering. They do all the design and decide on components that would have to be procured those that can be fabricated. But what was not visible was the old component in parallel. In procurement there will be two buyers.

One will be handling component A. Another guy will go ahead and take orders for this, the other components. The management had decided one month down the line we're going to cut out the old components. That information was not available, so the guy goes and orders for one year.

He gets a very attractive bulk discount and orders it. In the shop floor, there's again different tools development. There's another guy who buys the raw material and initiates the production for this for a batch of six weeks.

Now the General Manager makes the call, "This is the date. Let's go for it." Stunned, the guy who projected tells me this is the kind of stock, what do you want to do? It has been ordered, it's committed, and going to be delivered in a week's time, already the work is in progress.

Either you have to scrap it or go ahead and finish it. On the engineering side, I have gone ahead and finished the job. Marketing folks call the shop and tell that if they don't finish it in one month we will lose market share, it's a very painful decision, to scrap it, its actual case history, because information is not integrated and made available. But had this been made known in the planning process itself, then at least the guy ordering knows that this component is coming in, and this is the replacement of a particular component. Globally they have agreed that they will spend higher amount on this area in 2007. I understand that it's going to exceed even the ERP spending. Especially in the manufacturing side that's a great opportunity that requires a combination of skills. The manufacturing process from that perspective, in the IT and the technical skills, that's where we are.

So, implementing will be similar to an SAP implementation where... ... .

No, the fact, is that there are readymade products available, if a customer chooses SAP, Satyam has got about 3,000 people willing to go and implement it. Likewise you'll have people who will first advise the client about the process, where they can remove redundant process and bring in new ones and then help in implementation of already available packaged products.

There are two things, one is the package will be from the alliance with the vendor, but what we will plan is the integration so whatever the code, integrating and developing the interface is the value add. Maximum spending will be in the development of the interface, which will be in fresh code.

Developing the interface meaning connecting the SAP to those systems?

Correct.

From the ERP it provides the status of the manufacturing process itself, where the reflection is happening, where it needs to be weakened, what is the status of various products, where is the bottleneck in production planning happening, can I comment this time on for delivering, so that's what the manufacturing intelligence solutions.

There are also other areas, such as safety, manufacturing process optimisation itself, it's a practice itself for us. The chief of this practice joined us from Ford; she has been to Japan and Tokyo for production processes. She is building a team and that's going to be a key opening.

How do you think revenues from such integration would change the revenue mix for Satyam?

I think very soon the conventional IT will come down. For example when I came in, the Satyam enterprise in 1998, the revenue was almost close to zero when I joined.

Today, 40 per cent of our revenue comes from enterprise, and the engineering was almost nil but now it's more than 7 per cent. It will grow in a very significant way.

But, what is really interesting, which will put India on top of the map for manufacturing, I'm challenging the product itself. The product itself is becoming more intelligent. Take the mobile phone. It has a million lines of code in it. No one know this. So, who develops that? If you take the typical automobile, 23 per cent of the manufacturing cost is pure IT code. They say future automobile is going to be a computer with four wheels on the road.

That's what is clearly happening, it's completely and totally intelligent, be it safety navigation or entertainment.

The same thing is happening in aircraft, it's all high-tech. The same in mobile and consumer electronics and appliances. So everything is becoming intelligent and also communicating.

That's one dimension and the third is it's indeed a differentiator, I used to be a loyal user of Avis, when I rent a car, but I switched over to Hertz because of the navigation system.

They provide the `never lost' feature. You just key in where you need to go and it will take you through, that's the differentiator, even the cost they projected is a differentiator.

GM introduced an in-car probe. When you're driving, it keeps track — through the Wi-fi and wireless communication — of the status of various parts and issues alerts for those going to fail. There is a central room that communicates with this car. So it knows where the car is and then he directly calls the driver through the call centre and tells him your car is going to fail, so pull over. Also, the parts have already been dispatched, because the centre knows where the car is.

So that's going to be big differentiator, people enabled by IT to push the product.

So, because of that if you see, the automobile companies are spending 4-5 per centof the revenue on R&D, 8 percent of that is spent only on this area which is software-related. So all the differentiator is IT, so who makes that one time and who is going to maintain it ongoing, so who is going to be the automobile component vendor? So it's not an IT enabler anymore.

So you see that getting widespread 10 years from now? Because the software person gets only five percent of the revenue whereas 95 per cent goes to the hardware guys, maybe because of the capital intensive nature of the project. Tying up with such people makes sense.

\We see that already beginning to happen. We have recruited a person who has been in ISRO, he has directly worked for Abdul Kalam (the President, Dr A.P.J. Abdul Kalam). So while 30 per cent of the whole thing is software and 70 per cent is the hardware, why allow the customer to divide and rule? We'll be a single-point contact, this is software-tuned, that's what I said we reversed it.

We will go and talk to the hardware vendor and we may not want to get into the hardware part itself, but we could be the sourcing point. Instead of just recommending the hardware vendor to the client, we can take the responsibility of delivering the integrated solution. That's what all the aircraft companies are asking us to scale up and do. Why should they have to manage multiple vendors?

You are talking to people in India or... ?

India, in fact I was talking to a couple of senior people from big aircraft companies. What was attractive to them is the so-called offset, because if they sell aircraft, 30 per cent of that they have to spend here in India and that too only on engineering and manufacturing.

Initially they saw it best to include IT into that, but the government did not include engineering and design. So that's how we come into the picture. They want us to be a single point of contact and do the sourcing.

So you embed the software and give it integrated?

Yes. But now, they are a looking at India as a hub for the global market, even from a hardware viewpoint.

The other point they shared was, the kind of code that goes into the product is a very highly competitive one.

India is clearly seen in better light than China for protection of intellectual property.

That encourages us to develop that kind of system integration skills because that gives us speed.

So the interesting scenario is who's going to do what?

Whether the hardware vendor will take the initiative and acquire the software capability and or the software vendor will take it and integrate and deliver that end to end or would it be a third type of industry itself.

It's got a huge potential, what we are looking at is close to a $100-billion opportunity, rough estimates, that is.

India is rightly positioned, even ahead of China in terms of IT capability, and IP protection, the confidence level we can give, that's the bigger market share that goes beyond the commodity and maintenance services.

China has been far ahead in the no-brainer kind of thing, commoditised offerings. And, their infrastructure is better.

China is also smart. They will go in for backward integration.

Design is hardcore IP-related, and vendors may not like to give it to China. They would rather work with us Indians. When they give anything to China everything has to be finalised and even if they get the design, it would take them three to six months time to be able to replicate that. But vendors might not involve them in the product innovation stage itself.

So for any product launch it takes anywhere up to 36 months, now they are trying to shrink it to 12 months. We are involved from day one so they have a minimum of 12 months lead so that's the time it gets exposed to the Chinese manufacturer

So did you interact with hardware manufacturers, component manufacturers, and to select people whom to partner India with?

Yes there a few. It's in pockets... where we not only did the design and the prototype we got the product manufactured here with the suppliers, it started happening in a more ad hoc way

India is keen to grow the manufacturing industry as well. We have to leverage IT and also the branding goes with it. We are strong in design, engineering, software and IT, leverage that to manufacturing the product, not only from the cost perspective bur also for the speed. This could be an attractive perspective to the global customer.

So IT becomes a significant enabler for enhancing the efficiency of the internal manufacturing.

You have earlier talked about uniform value contracts. Just because you lock the customer into a certain fixed value, innovate and get productivity, does not mean that you can lock the customer in. Isn't their frustration over long-term contracts the reason for old contracts coming up for sudden renewal?

There are two components to that. One is pure vanilla outsourcing, like I'm using vendor X, he's doing all the maintenance enhancements, then can I have a vendor B coming in and take over. If both of them do offshore, then there's no rationale for the change.

In Europe that's what is or was happening - switching from someone who was fully onsite to someone who also gave an offshore flavour. They were kind of tied with the big five or the local vendors who did not have this kind of offshore flavour built into them so when Indian service providers go in they say they will bring in the offshore flavour to their maintenance, that becomes a subtantial saving to them as projected.

But what is happening is, it's not just the transition of the vanilla offshoring. Providing the advantage of offshoring alone is no longer a differentiation but the value add, i.e, services that are bundled along with the offshoring is significant.

For a big auto customer, we studied their entire business process and the application and said there is a great scope to rationalise your process, rationalise your application and if you do that you'll have a substantial saving and you can retire some applications which you don't need. The utility of that application is close to zero. Why do you need it, rather than maintaining it, because we did a user survey and nobody goes through that report.

If Indian vendors can do this, so can the rest. What is the true value add?

It's the focus of the domain and the approach that we take, the innovative approach, otherwise they'll say why, I've been working with this guy and he's not told me that!

What will happen is, it doesn't result in the kind of savings you have projected, then we say we'll factor that in. Then we commit for the next five years. That's where the uniform pricing takes place and we bundle the solution as a turnkey option that will be given to you and we take the risk for providing that.

To them, here is a vendor who has done the homework, put the numbers, who has given the proposal, even in that context they were fair, that they offshored that to the other vendors.

How about competition?

What's interesting today is that the customer is not necessarily looking at the reduction in the IT cost. If you look at typical organisations, IT spending is only like 1-3 per cent in the manufacturing, or even in banking it's 5-7 per cent.But from the CIO perspective it's significant because it's his whole budget. Whatever savings is there is great and if we can save even 30 or 40 per cent it's great. Whereas looked at from CEO perspective or CFO perspective, if I say I can save 50 per cent, that is half a per cent for a five-year period.

What they are looking for is across two dimensions, how to optimise and save business costs.

This is the cost of running your supply chain, this is the cost of running your back office and maintaining your whole financial and processing (process level and cost).

What's significant is how much time, how many users it takes to operate the system in use, for example the maximum spending was done on the finance people who use this to do their invoicing. The application was so complex it took so much time and it was not centralised or integrated, so it was sheer duplication and they are spending more time in reentering the data for different applications.

We, many of the Indian IT service providers are done in pockets, stand alones, we can say that for consulting, or we'll say we'll deploy the application or we'll say we do a remote infrastructure, or we'll do a BPO, but If we combine all the four - consulting, IT services, remote infrastructure management or BPO - this is the cost and then we'll take the responsibility of reducing your whole business cost which will be substantial.

You are incurring today X million, moving forward we are willing to take that as X minus 10 percent and this project is built into that.

But does it mean, if the IT budget was say $100 million and you had 2 per cent of that, this was outside the IT budget because you went to the CFO?

Obviously, if I had gone directly to the business head for the BPO pitch, it may not be substantial. If I go to the IT and make and make an IT pitch it might not be substantial but if I combine both and go to the user combining the process efficiency, IT and the BPO, then he listens to us.

In fact in one case the CIO was nervous about asking his CFO to spend more on his project, independent of the IT. The CFO had already asked the CIO to slash spending. They said we will go with you and present it to the CFO.

When we presented this entire picture, they cut the cheque and I became a champion for this project.

But we need to have the right people to speak the right language to the board and the CFO.

It happens in multiple areas, for example in business areas such as dealer management, auto financing...

We are picking up 12 areas where we want to play this kind of role, where we can take the entire responsibility of the business process and run it.

IT spending in manufacturing has not been very intense. When we recently met an auto CIO, he said, "I'll be happy if my budget stays static and stable for this year."

That's why we have more pressure to go beyond conventional IT. IT is not strategic for manufacturing whereas for banking and telecom, they can say its strategic, so that's why we have to get into more engineering, they don't mind spending more for engineering, the speed lies in engineering, and PLM, they don't mind paying more if they get the right skill and if you can give business value, so that they can hit the market with a higher, better product, faster than the competition, that's value add.

The engineering guy has got a better budget than the CIO.

Now the shop floor, that's again the strategic point, rather than just doing the IT on the business side, then it's a savings, then you have a higher growth and higher margin.

bharatk@thehindu.co.in

More Stories on : Interview | Software | Satyam Computer Services Ltd

Article E-Mail :: Comment :: Syndication :: Printer Friendly Page

Stories in this Section
Find your way

Site access problem

`Time to retire Telegraph Act'

Linking the shop floor to the top floor...

Hackers on the prowl

Can't beat `em? Join `em!

The `BSC' of identity

Quiz

A cardinal sin called overconfidence

Cartoon