Employees and customers alike could be forgiven for being confused when Ian deSouza became president of Universal Instruments Corp. last August. After all, he not only shared a first name with the person he superseded - Ian McEvoy - but the two have been close friends for years.
 
What is not confusing, however, is deSouza's approach to running the placement equipment maker. As senior vice president of operations he had been on the executive team for years and played a large role in formulating the company's strategy, first for recovery from the recession and now for a new industry growth phase.
 
An avid amateur pilot - he says his first solo came when he was "11 or 12" - deSouza spoke with Circuits Assembly's editor-in-chief Mike Buetow on Asia's maturity, the need for tighter development windows and the pros of decentralization. Excerpts.
 
 

CA: Where do you see Universal these days? Are you aiming for a tech leadership role, or positioning as quick follower up the ladder?
ID: I think we have a technology leadership role - we have for many years - and we're not going to relinquish that without a real fight. I don't see us doing so, for a host of reasons. We've maintained a position that is important for us, not the least of which is that we've reinvested 20% of our annual revenues in research and development. We have done so in tough times and good times. That's part of our culture.
 
CA: Have you benchmarked that against competitors?
ID: We'll benchmark it against anyone you want to benchmark it against.
 
CA: That's at least twice what a typical electronics company would contribute.
ID: Probably. We've been a dominant player since the 1980s. We were the first to come to market with the platform the industry recognizes as the GSM. It became part of our structure - our thinking. The latest products, which are really now two years old, have a lot of life to go.
 
We have better than 40% market share in the flexible arena. The reason we do is because there's a need for that combination of versatility with the raw chipshooter to get the throughput. We recognized that as the chipshooters became broader, we had to become broader ourselves, to protect our position, leverage that position and increase our market share. So by introducing an ability to handle chips at high speed on that platform while retaining all the good, positive qualities of being able to handle complex components, a wide component range, etc., we would egress up the chain. That's really our mission for the next few years.
 
Our egress is happening. We're gaining market share. It's happening more slowly than I anticipated, but when I think about what we are trying to do, the pace is appropriate.
 
More and more customers are recognizing that the Lightning head in combination with the capabilities that we had on the platform itself really does allow them to increase the capabilities of their lines. We're moving up within our own customer set and gaining share by winning customers from our competitors as that momentum continues.
 
CA: When you look at share, what do you count?
ID: We look at three things. Market share as it pertains to the whole industry - screen print through final assembly. We look in terms of units, and in dollars.
 
Our primary tools are industrial bodies. One body that gives us the majority of that data in Japan is Protec; they collect good data. We moderate that with a combination of data collected from publications, our own information and every two years we do a fairly extensive market analysis, sponsored by Dover, which includes market share data.
 
CA: How tightly intertwined is Universal to Dover?
ID: Dover is the financial institution - I use this analogy internally frequently - and this is our business. Dover doesn't interfere with how I run this business. Why is that important? It allows us to flex and turn and modify our behavior in accordance with what the market requires.
 
When I need to talk with somebody, I'll go and talk to them. If I need a market study, they'll be glad to sponsor it and be part of that process. But they are pretty much the banker.
 
CA: How is Dover's model different from competitors who aim for vertical integration?
ID: At the end of the day, there is a critical mass at which companies become slow and cumbersome. And combining different companies with different cultures always makes that more difficult, although there are good organizations that have proved that can be done and done effectively. The Dover philosophy - and I really believe in it - is that you have to allow your companies to be versatile and have the passion themselves to grow and understand their business, and you only do that if the organization is versatile enough to do that. Dover at different times has asked, "Would it be better to put DEK and Vitronics and Universal together and have this critical mass that's that much larger?" Each time we test it we come back and say, "No, it doesn't make sense." It's best to allow each of the pieces to act independently when it's in their best interest.
 
Vitronics is number one in their market, DEK is first in theirs. I would love to be number one, and that's where we're heading. And I can do whatever I need to do, unencumbered by the other two. At the same time, when it makes sense for the three companies to work together - and we have a laboratory here in which we've invested millions of dollars, and Dover made that investment - that's when we work together.
 
CA: How often do you meet with your counterparts at Vitronics and DEK?
ID: Two or three times a year. And we talk on the phone more regularly, especially if there's something we want to work on.
 
CA: Do your engineers have carte blanche to commingle with their Dover counterparts?
ID: If they want to, they can pick up the phone and talk to them. And that's one of the bizarre things; it really is like talking to somebody at another company. There are some ties there, and you know who to call, but that's about it.
 
CA: How do the labs prioritize in terms of customer applications and R&D?
ID: The labs are kept independent. There are two activities the labs have. One is called consortia. In that arena, we're just the host. The consortia include a lot of big name companies. The task is to look five years ahead, to begin to investigate and understand the equipment and processes to accommodate them. There have been major adoptions in the industry that were allowed to happen because work was done by the consortia in our New York lab. The second activity is independent, but uses some of the knowledge that Universal gleans from that (or DEK, etc., and its process support for customers.) We do everything from failure mode analysis through to development of new processes, and research specific to Universal's direct customers, which is a unique advantage of being a Universal customer.
 
CA: Why five years ahead? Why not three or seven?
ID: Because I think that that's as long as you can look ahead and get it right. Any shorter and you don't have time to respond; any longer and you're playing in never-never land.
 
CA: How much customer input is taken into account?
IM: It's more of an industrial roadmap. We don't make decisions on our product based just on what comes out of the consortia. The consortia quote often is one layer ahead of where the equipment is; it's the big cellphone manufacturers, the big automotive guys saying, "What do we need to do at a macro level in terms of processing and development?" We tap off that for intelligence and information but we look in other places too.
 
CA: How much of what your lab is working on is driven by what your customers are asking for?
ID: All of it.
 
CA: So there's no mad scientist in the bowels saying, "This is what my vision for pick-and-place is ..."
ID: Actually, we have two mad scientists on the consortia that have very strong personalities and are well respected. There are also some pretty clever mad scientists on the customer teams and they participate in the decisions regarding the direction of the lab.
 
CA: How heavily are your sales engineers involved in collecting customer feedback for the labs?
ID: They aren't involved in the process. The consortia and the lab activity are not our development activity. The consortia bring customers to the process side of that organization, and they will sell the consortia to the customers. The consortia develops processes and pushes equipment and materials manufacturers to address things it finds in its analysis, etc., that have to be addressed for these processes to work in three to five years. In terms of product development, we have regular forums in which various bodies from the organization, including sales, get an opportunity to talk to the right people about what they are seeing and hearing and that gets put into the intelligence.
 
CA: You have spoken of the need for North America to push up the technology ladder. In your mind, is technology in and of itself - the ability to do something no one else can do - a solution for profitability or just a means to an end?
ID: It's not technology for technology's sake. Absolutely not. I'm an operations guy. The simpler it is, the happier I am. Sometimes you need technology to make it simple. What North America needs to do, by default, is to look at ways to make it more effective to manufacture here. And it's also the place where processes will be developed, primarily because the new designs are developed here, and you want the manufacturing processes and the design to go hand-in-hand. Having said that, as China and other places become more mature in terms of their own design and the end product, you're going to see the need to do similar things with the development of processes there, which is why we opened the lab in China. That need is already there, it's picking up pace, and North America is already in competition with them.
 
CA: Do you think North America and Europe need to adopt giant industrial parks like the one in Suzhou?
ID: Generic answer: Does it help to be near your customer and your supplier? Yes. Is it well up the totem pole of the things you have to do as a business to be successful? I think there are many, many other things that give you far more power than that does. Design, understanding the market, creativity, business strength. There's a host of things that constitute a business. Universal could be here or China - its power is how we run the business. ... Singapore has had industrial parks for years ... and there are cells in the U.S. of course. ... The China model required a lot of gains very quickly to be competitive because they were so far behind other players. And kudos to them for recognizing that you can only do that in a combined, focused way.
 
CA: You worked with [former Universal president] Ian McEvoy for years. Is it strange not to have him a few doors down?
ID: I wouldn't say it's strange. We talk - he and I have been best friends for years - as we did when I was in a different company, sometimes about work, sometimes not.
 
CA: Is it easier to take over a company that is doing well than one in turmoil?
ID: This is the first time I've taken over a company so I don't know what's easier. [Laughs] I don't get fazed by a lot of things. I don't intrinsically panic. I think it's my nature. I mention that because my focus wasn't on me becoming president, but what on the company is doing, and whether we had good plans.
 
CA: Because of the way and frequency that they relocate, and with so much of their product cost now tied up in shipping, it seems top tier CEMs are as much in the logistics business as they are in product development and making solder joints. How does that translate to what you do?
ID: It's an interesting question. Their product lifecycles and design cycles are much, much shorter than ours. So their opportunities to move are much greater than ours, because I think you have to move it across at intrinsic points in the product; you can't do it haphazardly through the life of the product. We have a facility in China; a large software operation in India; and consortia in the U.S. and in Suzhou. Beyond that infrastructure, we have this core infrastructure; like Dover, a very decentralized structure. That's our base. We expand and contract the customer support functions along with our customers, but our own core infrastructure of design and manufacturing moves very slowly.
 
CA: When a customer with 100 factories around the world suddenly decides to shutter 30 of them, what do you do?
ID: There was a lot of capacity that was put in place in 1999 and 2000. The reality is that the equipment that was put in place then, I think, has about eight years of life. The curve has started to drop off very quickly in terms of the cost to maintain it - and to a point it doesn't matter whose sets you are looking at (although I do think the GSMs are proving more robust, because they were the only gantry systems at the time). It will be interesting to see what happens in the next two or three years with all this capacity. We're already seeing customers compare the cost to maintain versus buying a new machine, particularly as the new machines' prices have come down so substantially. Maybe it's time to change, and maybe it will create a little euphoria.
 
CA: Do various depreciation laws play into this at all?
ID: I don't think so.
 
CA: What type of lead times do you maintain?
ID: We try to keep them between two and eight weeks. Everyone wants to condense the machine lead times. There's a regional difference, but you can see the maturity in Asia, which is becoming more realistic. It has evolved and there are places I think are more sophisticated than the West in their practices.
 
CA: Such as?
ID: I think their realism about what's important in equipment and in their manufacturing operations; their consciousness around cost; the value they are getting out of things.
 
CA: Some want to write off North America as a place where volume manufacturing will no longer be done. Where do you stand?
ID: I think for the foreseeable future, things aren't going to change dramatically. China's going to be the volume manufacturing center. India's going to develop a lot of software. Central Europe will supply Europe to a degree, but probably won't compete with China. The Americas will continue to evolve and develop technology, introduce it and look to move it to another place. Over the last 12 months or so, we've seen a comeback in Mexico.
 
CA: What's the window for Universal for product development today?
ID: It's getting shorter, and the need for me to get it even shorter just grows. It's a place of focus. Ten years ago it took five years to develop an entirely new machine. Today it's three years, and we'd like to see it come down further than that. People will tell you they do it in 15 months; they don't.

• Prev
• Next