Partnership Model, Open Innovation
Leading PARC into New Era of AI,
Bridging Research, Business Value
Dr. Tolga Kurtoglu is CEO of PARC, a Xerox company, which is in “the business of breakthroughs.” Practicing open innovation since being incorporated in 2002, PARC provides custom R&D services, technology, specialized expertise, best practices, and intellectual property to Xerox’s business groups, Fortune 500 and Global 1000 companies, startups, and government. Prior to PARC, he was a researcher at NASA Ames Research Center, and a mechanical design engineer at Dell Corporation. He received his Ph.D. from the University of Texas at Austin and M.S. from Carnegie Mellon University, both in Mechanical Engineering. He holds a bachelor’s degree in the same field from Or
ta Dogu Technical University (ODTU), Ankara, Turkey. He recently took a few minutes to talk to AI Trends Editor John P. Desmond.
Q. Can you describe the mission of PARC and what your priorities are in the near term for our readers?
A. It’s been about 18 months since I took over as CEO of PARC. I was familiar with the organization prior to my appointment as CEO. I’m thrilled to take on the responsibility to lead such a renowned and legendary organization. The mission of PARC is essentially to create the open innovation center of the 21st century. We’re in the business of breakthrough innovation. That requires the ability to translate science into market impact. And that’s what we’ve been doing for many years.
Q. The Xerox Palo Alto Research Center has a rich history as the birthplace of personal computing and the graphical user interface, and founding generation of engineers that have gone on to create their own companies, including Adobe and 3Com, and inspire new companies such as Apple and Microsoft. What’s it like to be the leader of an organization with that legacy?
A. It’s fascinating. PARC always had some of the most brilliant scientists and engineers under its roof, which led to a lot of those early inventions and the creation of new industries and technologies. And that is certainly true today. We continue to have some of the best and most brilliant scientist today coming from various different fields.
PARC has always been about the future. When it was first founded, it was about the Office of the Future and a lot of those innovations came about in that context. We continue to envision the future and bring technologies to bear through science and technical investigation to help our clients and partners share that vision going forward. A major difference is that we have adapted in a completely new business model today, with a growing number of partners and clients today.
Q. How does the mission of PARC today fit in with the mission of the parent company, Xerox?
A. We continue to represent the innovative face of Xerox today. We work very closely with the corporate strategy and the product teams of Xerox Corporation to continue serving their R&D and innovation needs. I will mention two areas of focus. First, we continue to think about our capabilities in print, and how to think about print in radically different ways that are beyond the paper and beyond the ink on the paper.
Second, a big part of Xerox’s business is about information flow and sharing of data and information and connecting that in an enterprise setting. We have a number of technologies at the intersection of artificial intelligence and augmented reality, and come other emerging technologies. We to continue to innovate on multiple fronts for Xerox Corporation.
Q. In your previous position as Director of the System Sciences Lab at PARC, you were responsible for research areas including AI, machine learning, modeling and simulation, cyber and physical security. What is the focus going forward for this AI and related research at PARC?
A. A lot of that stems from observations of where the world is heading. We are seeing more and closer integration between the cyber world and the digital world. Many new new machines and systems are examples of that collision of the cyber in and the physical world. It may be the self-driving car, your Nest thermostat, your autonomous cleaning robot, and so on. We see many examples of where we need to think about the world at the intersection of hardware and software, the cyber and the physical. We have quite a bit of work going on in how to apply AI, machine learning, modeling and simulation to systems of the future.
We see three major research focus areas going forward at PARC. One is how to enable future systems to sense the world around them in a better way. When we talk about IoT, we talk about the abundance of real time data. However, we need access to the right kind of data. At PARC, we’re building some novel sensor capabilities driven by our capabilities in producing low cost electronics, including printed sensors and printed electronics, radically new ways of sensing the world. That is a major focus going forward.
Another focus is elevating machine intelligence. Once you have the right kind of data, you need to elevate the intelligence of the surrounding machines and systems in a way that they can actually perceive the world around them in a much better way. We are working on applying AI to for example computer vision flying, in order to elevate the level of intelligence applied to processing the data coming out of these systems.
A third are of focus is making the more highly-connected systems, be they self-driving cars, drones or robots, more secure and safe. We are thinking about and researching reliable and trustworthy operations of these systems in the cyber physical world.
Q. Could you describe the partnership model at PARC and how it delivers on your mission?
A. The PARC system is really at the boundary of R&D. It’s all about translating deep science and technological investigation into commercially viable pathways for productization. That translation is not very easy, and that’s where we sit. And we do collaborative innovation with an ever growing number of partners. We provide some of the strategic innovation capabilities. We have a workforce that really understands emerging scientific trends and technologies. We work on embedding technologies coming from our ideas and inventions into the business context of our clients and partners. We work closely with the internal R&D and new venture creation departments of our partners. They bring a great deal of domain expertise to the table, for strategic business problems that require technical and scientific solutions, and we bring our collaborative working principles, our innovation culture, as well as our creative ideas to actually build solutions. We help to build tangible prototypes and proofs of concept. I think of us as an extended R&D node of all the partners that we’re working with.
Q. Can you highlight any projects that are good examples of how the partnerships work?
A. One from the area of machine intelligence is with our strategic partner East Japan Railway Co. (based in Tokyo) which basically operates the train systems.
They specialize in designing, manufacturing and operating the train systems. They’ve been collecting significant amounts of data with respect to their operations. They wanted to improve their ability to approach maintenance and operations in a much more intelligent way. That requires quite a bit of expertise in computer vision, machine learning, AI, and data science, which is what we have provided. We work with them very, very closely to understand their operations and look at the real time data coming out of their train system.
We started with train doors, then we moved into switch machines, then the rail tracks themselves over the course of a couple of years. We built a platform which they have operationalized, that now helps their service and maintenance department look at the real time data and make predictions about when components are likely to fail and why. That enables them to deploy their maintenance and service resources in a much more efficient way.
Another example is with partner Procter & Gamble, which had a different kind of problem. They wanted one of their cosmetic beauty care brands to stay connected with their consumers and their target market segments. They had a disconnect between their traditional way of engaging their target clients versus what the emerging generation is more accustomed to. Our human-centered design and user studies group suggested a very different kind of experience with technologies developed at PARCK embedded into it.
We studied the behavior of the target consumers and helped our client build a mobile application platform which enabled interaction at the level of accuracy that you would get from an expert cosmetologista skin care expert. We created an app with them that can analyze the skin characteristics of somebody, come up with a detailed analysis, then make some personalized, scientifically grounded product recommendations.
Q. Changing gears, who would you cite as a competitor to PARC?
A. Our biggest competitors today are startups. The center of gravity in the innovation ecosystem is in my opinion moving to the startups. They operate in a very, very agile fashion, and effectively as well in a lot of ways to bring new technology to the market. And since our ultimate goal is also to bring technology to the market, how to translate science into market impact, we need to be very aware of the startups. We need to bring our technologies and ideas to the market as fast and as agile as the startups.
Q. Do you still have pure research going on at PARC or is it now more now focused on practical applications?
A. Absolutely. We have a balanced portfolio between basic research and practical applications. We do foundational research in for example developing new sensing capabilities, and we do quite a bit of work thinking about applications spaces, domains and uses cases of those technologies. We are really good at bridging the two. There are many challenges to bridging basic research to the right set of applications that can unlock value in the marketplace.
Q. In July 2017, the Defense Advanced Research Projects Agency (DARPA) selected PARC to help develop an “explainable AI” system called Common Ground Learning and Explanation (COGLE). How is this work going?
A. That work is going great. We had some early demonstrations of the technology we’ve been developing very recently. I’m excited about the project because I believe it addresses a very fundamental problem with the adoption of AI technologies, which are penetrating the consumer market and enterprise space. We set out to explore how to break the black box nature of the AI algorithms. Many modern AI technologies, such as deep learning, operate in sort of a black box fashion. You feed in a lot of data and you let the algorithm essentially find patterns in the data, learn from it and come up with recommendations or decisions. But if you ask these algorithms why they come up with a particular recommendation or decision, they’re not really able to articulate themselves. And that’s really creates a whole set of problems when it comes to accelerated adoption of these algorithms.
So we’re trying to bring some transparency into the inner workings of these algorithms. In a nutshell, what we’re trying to create a translation between how the human brain represents information and how these algorithms represent information in a semantic fashion. It’s a highly interdisciplinary research. We are working with some of the best cognitive scientists in the world on it, and we’re excited about some of the early results of the project.
Q. You worked as a researcher at NASA Ames Research Center. Last December, PARC announced a partnership with Blue Origin, Jeff Bezos’ spaceflight services company, to conduct R&D and surveys. Can you describe this work?
A. When it comes to R&D in space, we’ve had development projects in the past with NASA, using our printed electronics and low-cost electronics capabilities. That lead Blue Origin to think about what kind of research we can do in space together. It’s an ongoing process. We’re still trying to define the exact scope of what we will take on.
Q. The US Department of Energy in November 2017 announced a partnership with PARC to help increase the reliability and resiliency of the electric power grid using smart monitoring. How is this work going?
A. That work is going quite well also. We wanted to help develop technology that would increase the security and the reliability of the power grid. We have a long standing set of technologies that we’ve developed for smart monitoring at the intersection of sensing and analytics, and machine learning and AI. In this case, we applied it to transformers, which is one of the weakest links in the entire electric power grid. At the heart of the technology is a fiber optic sensing technology, which is fairly mature. We’ve developed some analytics and smart monitoring capabilities using that core platform in the past for electric vehicle lithium-ion batteries. We’ve also deployed the same sensing capabilities for structural monitoring of critical infrastructure like bridges and highways.
When we saw this part of the problem with the grid, we got really excited and we thought there was an opportunity to essentially customize and apply our fiber optic sensing technology to a new problem. That’s precisely what we’re exploring now. We have developed the appropriate AI and data analytics capabilities that enables us to process sets of data coming from the sensors in real time. That’s very important when it comes to monitoring these transformer units in the grid. So it’s a relatively new startup project but we’re developing in applying and fairly mature technology to it, and we’re really excited about the potential impact of that.
Q. Is there anything you would like to add or emphasize?
A. I would emphasize that the partnership model and the open innovation model are clearly unique to PARC.
Learn more at PARC.