James Spohrer: How systems interact to deliver services

December 02, 2009

By the time you reach your office every day you have already tapped into numbers of service systems. The electric grid that powers your alarm clock, lights and coffee pot, the pipes that deliver water for your shower and the transportation network for your commute are all service systems -- networks of systems actually. Service science is an emerging discipline aimed at understanding how these systems interact to co-create value. The IBM Service Science Initiative has led the way in promoting research in this area. Dr. James C. Spohrer is director of service research and innovation champion at IBM's Almaden Research Center. He was a featured speaker at the Compete Through Service Symposium, presented by the Center for Services Leadership at Arizona State University. In a telephone interview recently, Dr. Spohrer talked about the nature of service systems and the research that is expanding their potential to improve the way we live. Podcast coverage of the Compete Through Service Symposium is sponsored by IBM. 15:07 



By the time you reach your office every day, you have already tapped into numbers of service systems. The electric grid that powers your alarm clock, lights and coffee pot, the pipes that deliver water for your shower and the transportation network for your commute are all examples of service systems. They're networks of systems actually. Service science is an emerging discipline aimed at understanding how these systems interact to co-create value. The IBM Service Science Initiative has led the way in promoting research in this area. Dr. James C. Spohrer is Director of Service Research, an innovation champion, at IBM's Almaden Research Center. In a telephone interview recently, Dr. Spohrer talked about the nature of service systems and the research that is expanding their potential to improve the way we live.

James Spohrer: Service science -- or service science management engineering and design, as it's sometimes called -- is an integrative field that studies different types of service systems in the world. Service systems are all around us, and they're certainly understudied. Every day we're customers of about 40 different types of service systems in the world: from the minute we wake up and turn on the lights, the electric service; to when we drive or take the bus to work, transportation services; to when we pick up our morning coffee, our retail service; there's just a huge number of service systems that we depend on for our quality of life these days. For a variety of reasons, a lot of the focus in university disciplines and research has been on product systems and not so much on service systems, but it seems like that's beginning to change.

Knowledge: Okay, so this is a science that studies systems. That would suggest that there is an interdisciplinary aspect to it?

Spohrer: Absolutely. Service systems typically are configurations of resources -- like there's people involved, there's technology involved. There are usually other organizations that provide some of the capabilities, and there's information. Every service system has to try to configure those resources to help co-create value with customers.

In service systems the customer's role, because the customer brings resources to bear, is very important, and that's really one of the main things that differentiate service systems from a lot of the factory and manufacturing systems of the past. Sometimes that role may seem pretty minimal: like if you're getting a haircut, it's primarily your time, but oftentimes, it's a lot more significant. So if you think of a medical or healthcare service, the quality of the information that the customer provides the doctor can have a profound impact on the quality of the service that the doctor is able to provide, and the role of students in an education service is pretty profound as well.

Knowledge: Well, it seems to me that there's an intersection here between, maybe, social science and what we might call hard science or the technological sciences.

Spohrer: Yes, there's tremendous complexity in it because service systems have people. They have technology. They have other organizational relationships. And so these are some of the most complex types of systems that exist on the planet. We're really trying to focus on this phenomenon of value co-creation when different types of systems, with different types of resources, interact and try to co-create value. I would say two of the most important types of service systems that we work on at IBM are other businesses and transforming other businesses. Increasingly what we're seeing is that more and more of our business services from IBM are really in the category of B to G, or business to government.

The service systems to think about -- [for example] there is a city, [and] cities are very complex service systems. In fact, you might even say [cities are] systems of systems. When we do transportation projects, we'll work with a city to make sure that all of the major arteries that lead into the city, have cameras that can look at the license plates of the vehicles going in during the rush hour.

We did a project in Washington State where, when consumers were about to use their dishwasher or their washing machine, they could see information about what the cost of electricity was at different times of the day. People start changing their behavior when they understand what the cost is of using different types of service at different times. What we're currently working on is developing better and better models of these systems. Ultimately we'll have a computer-aided design system that will help us better manage and engineer these complex systems, but right now that's kind of a Holy Grail.

Knowledge: So this is -- I think everybody would agree -- a fairly young field, at least the way we're studying it now. Have there been any significant milestones that have come up in the course of this research?

Spohrer: Yes, there have been a number of things that are, I think, quite significant. When we started the service science effort, about six years ago here at the Almaden Research Center, it was clear to us that there had been a lot of very, very good work in a very few places in the world; but it was also clear to us that it was somewhat siloed and not integrated. One of the big challenges was to get an integrated initiative.

There's actually over 400 universities in the world now that are teaching service science-related programs in an integrated way. We're seeing a lot of different academic disciplines starting service science segments as a result of our effort. I guess there's really only two things I regret. One is, I don't know why the world didn't start sooner on service science and service research. When I think of all the benefits that we could get if we had just started sooner, but except for a few real pioneer institutions, like Arizona State University, that really got in there early, over 30 years ago, it really hadn't hit the main stream, I don't think, until service science initiative started six years ago.

One other kind of proof point for the success of this is that we, in the Almaden Service Research Group, achieved an eight times return on investment after six years; $290 million worth of impact on IBM's business in just six years. Good, good validation that if we took a systematic approach to improving service systems, there was a lot that could be done.

Knowledge: You mentioned that developing a model for service systems was the Holy Grail of services research. What are one or two other compelling questions that could be worked out at this point in time?

Spohrer: Once you have really good models, then you really want to make sure that you have good data to go along with those models, so more and more service systems are getting sensors. Even underground now there's a lot of sensors at the traffic stops; there's just a huge amount of sensors that are being put into the world. This goes along with our Smarter Planet initiative where the world is becoming more instrumented, and it's becoming more "intelligent." We're able to store and process that data about these service systems. Once you have the models, then you want to keep the data up to date and, and really, do a lot of analytics, looking for insight.

I was just talking to someone who was telling me about when you have the appropriate traffic data, you can really start identifying information about where the traffic, where the accidents, are most likely to happen -- and start saving lives. When you have more of the data about healthcare you can gather insights about what treatments are most effective. The more we can instrument these service systems, and gather the appropriate data, and do analytics on them to get insights, [the more] we can start improving them.

The third one is really trying to discover a Moore's Law of service system improvement. We should be able to put service systems on learning curves so that we're using that data to improve productivity. We're using it to improve quality, compliance and really make innovation more sustainable.

Knowledge: You know, I was thinking that from your chair at a research institute at IBM, you're in a perfect position to see what's coming up in terms of improvements in service systems. I was wondering if you could tell us what might change that might be really surprising for people in the next couple of years or even further off over the horizon?

Spohrer: Well, we have a lot of interesting research projects, and a lot of them are in collaboration with our university partners. We have a lot of efforts around cloud computing -- this is about computing installations that allow researchers from around the world and organizations from around the world, to really look at massive amounts of data and then do computation on them.

Cloud computing is really going to transform, I think, a lot of the way science is done. The data is going to be collected into the cloud; and the processing is going to be relatively low-cost, to do massive types of analytics over massive amounts of data at a scale that I think is certainly going to just transform the way that science is done; and should lead to an accelerating rate of discoveries in many different sciences. I think that's really going to be an enabling technology for a lot of new types of service innovation. Some of those implications are pretty profound for science.

We also have a project -- again, with universities -- where many people know that IBM built a super computer, about twelve years ago, that beat the world champion at chess. We're currently working on a system to beat the world champion in Jeopardy which, for anybody who knows the game of Jeopardy, it's just the people who win Jeopardy are the -- like walking encyclopedias. We're building a system that will be able to answer natural language questions better than any person can do that. It's going to be primarily an example of how computers are getting more and more sophisticated: to have natural language, interactions with their users. Again, it's also an example of amassing massive amounts of data and making it accessible on a global scale.

We also have a project that's quite interesting that should complete around 2017 that's a very far-reaching project called Cognitive Computing. As advanced as our supercomputer technology is, it's not as advanced as, in some ways, our human brain is in processing information. To run the supercomputers -- the large scale ones, the giant ones that are available, for instance at Lawrence Livermore -- it takes whole dedicated power stations just to generate the energy required. [With] the Cognitive Computing research, we're really trying to figure out how to get computing power of the order and magnitude of the human brain, that's also low temperature and low energy. So that will be pretty transformative to have massive amounts of computing storage processing capability but not as expensive in terms of energy. All of these technological advances, of course, enable new types of service and service innovation.

Knowledge: You were talking about the fact that we now have the ability to be able to detect the license plates of cars that are traveling on a freeway during the rush hour, for example, and that perhaps we may, in the future, be able to develop a tolling system, or perhaps we may be able to redirect traffic to make things a little bit more efficient. I was wondering is there anything like that that might affect daily life that's coming out of this services research?

Spohrer: One of the things that we're doing to help make progress there is we've started something called the Smarter Scientist Initiative. People can go to www.ibm.com/think to find out more information about Smarter Planet. View the world as made up of a lot of different nations: there's usually a few big cities in those nations, and there's usually a few universities at the center of those cities. Oftentimes those universities are what are called urban-serving or region-serving universities, where 60 or 70 percent of the students are coming from the region, and 60 to 70 percent of the students' first jobs will be jobs in that region. So these universities are just incredibly important for creating and disseminating knowledge and training of the next generation of workers for those regions. From a service science perspective, those are all service systems.  The university is a service system, the city that it's in is a service system and the nation that it's in is a service system.

We've started building, what we call, component business models, CBM models, of all these different types of service systems -- what all their key component indicators are -- and really trying to do a much, much better job at understanding how do you invest in all of these systems to make them more sustainable; to improve service quality; to improve service productivity.

So I really believe that the science, the technology and engineering, the mathematics, the improvements in management of service systems, that we'll be seeing in the next decade, might make us want to call the next decade, the decade of service. Because there's just going to be so much potential to improve the service systems, that we're finely studying scientifically over the next decade, and the decade after that will end up being technology to improve these systems. That, again, I really think it's going to be a great opportunity for all of us who are interested in service research, to think about making this next decade the decade of service.