CTO Session: Gerardo Pardo-Castellote, RTI

Are there any technologies which you think are overhyped? Artificial intelligence and autonomous driving… expectations are likely much higher than what can be delivered in the short term, but eventually will be real and practical.”

Headshot of Gerardo Pardo-Castellote, CTO at RTI

Name: Gerardo Pardo-Castellote

Company: Real-Time Innovations (RTI)

Job title: CTO

Date started current role: January 1999

Location: Palo Alto, California

Gerardo Pardo is CTO at RTI, he has been with the company since it was founded, held various positions in research, development, consulting, and became the CTO in 1998. Pardo was the main developer of the original versions the RTI Connext DDS product line. He started this work while still working on his Ph.D. at Stanford and carried on the product at RTI. Pardo leads RTI's standardisation efforts and he is the main author of the OMG Data Distribution Service Standard (DDS), the OMG Real-Time Publish Subscribe (RTPS) Protocol standard, the DDS Security Standard, and many of the other (currently 7) DDS-related standards. He also chairs the OMG revision task forces for these standards. Pardo is considered an expert on security, middleware, and publish-subscribe systems. He has been the Principal Investigator on more than a dozen research grants, including SBIR and ATP grants, has authored five patents and over 50 technical papers, journal articles, and conference sessions on the topic of real-time and secure distributed systems.

What was your first job? I had some summer jobs before going to college. The first one I remember was in agriculture, picking almonds. It was a manual job back then.

My first career-oriented job was at CERN, a European Center for Particle and Nuclear Physics research. I worked there two summers while I was still studying my Undergraduate degree at the University of Granada (Spain). My work was developing software for modelling and control of various parts of one of the experiments there. It merged aspects of physics (my undergraduate major) and Computer Science.  

Did you always want to work in IT? I was always fascinated by computers. I got some of the first 8-bit “consumer-grade” computers when they came out in the 1980s. I was still in high school and the fact that you could get a machine to solve mathematical tasks completely fascinated me. I programmed it and used it to check my algebra homework which was hand-written on paper back then.

At the time I was on the high-school chess team. In our free time we tried to program the computer to play chess. It was considered a hard problem then, supposedly illustrating some of the unique aspects where humans were “better…”  Not for long, I remember when the first dedicated computers for chess came out, they had an integrated chessboard and it was easy to beat them. Now computers can beat any human at chess and any other board game.

I was inspired to work in robotics by all this. If you could combine the capabilities of computers to “think” with the mechanical advantage of machines it would be possible to automate any arduous human tasks and improve the life of many people. I guess it came down to that first job (picking almonds) not being so enjoyable :)

What was your education? Do you hold any certifications? What are they? I did my Undergraduate studies in Spain. I majored in Physics with a minor in Computer Science, both at the University on Granada.

I came to the USA for my Graduate studies. I did a Master’s in Computer Science and a Ph.D. in Electrical Engineering (Robotics)– both at Stanford University.

I have not pursued any formal certifications, but learning is something that should never stop in the technology field. I spend a lot of time reading and trying to keep up with the current technology developments.

Explain your career path. Did you take any detours? If so, discuss. I started my career in Physics. I like that it helps people to understand the real world. And the fact that it is grounded on experimentation and has “reality” as the ultimate arbiter of debate. This led to my summers at CERN and I was considering a career as a physicist there. However, at CERN I was introduced to Real-Time Computing.

High-energy physics experiments produce massive amounts of data. They look for extremely low probability events. The experiments produce too much data (and too fast) for storage and post-processing. Rather it must be examined in real-time and a decision made to discard the data or save it for a deeper off-line analysis. This is done by applying a pipeline of tests, starting with the simplest/fastest that can be used to discard most of the uninteresting data. The tests may combine both hardware and software approaches, such as timing the coordinated triggering of two sensors by placing a length of fibre-optic cable between them whose length (in terms of speed-of-light distance) corresponds to the expected delay.

I was impressed by the potential to use software to analyse/interact with systems in the real world and make decisions faster than any human could.  I also came to appreciate the importance of Software reliability and quality. Some of these experiments need to run for years before the desired event is found. Dependability is paramount.

It was a fascinating job. I got to interact and learn from many talented and truly motivated people that wanted to advance science and technology. It made me realise the power of software and inspired me to pursue a career in Robotics and Computer Science.

What type of CTO are you? I am an experimentalist at heart.

It is comparatively easy to have ideas. It is harder to discard them and select the most practical and those that will maximally impact your users and markets.  It is even harder to implement those ideas such that the resulting product is robust, maintainable, sufficiently intuitive, and meets the functional and non-functional requirements of the users.

All this is even more applicable when users are developing “real-world” systems that control critical infrastructure and have to be robust and highly available.

I am engaged with the architecture, design and implementation parts of our products. I am involved in the brainstorms about challenges and future directions as well as in the coding/prototyping of some new features or products.

I believe that the key to innovation and creating robust products lies in teamwork and collaboration. It is essential to surround yourself with talented and motivated people and foster an environment where people can develop their passions and where creativity, collaboration and hard work are cherished. I focus a lot of my efforts in building the team and fostering the culture.

Which emerging technology are you most excited about the prospect of? Networked Autonomous Systems. This goes beyond specific applications, such as, Autonomous Driving and Robotics. Rather it is the general combination of Machine Intelligence with complex Electrical, Mechanical and Industrial Systems to perform more coordinated and autonomous operations without requiring tedious human intervention.

We are seeing a disconnect between resource production and efficient use, resulting in significant wasted resources, be it in food, materials, or energy. Consequently, we are using natural resources in non-sustainable ways. 

I believe better integrated Autonomous Systems will be an important part of the solution. For example, looking at the electric grid, if we could control the production and distribution network in a more integrated manner, it would be possible to regulate peak demand to a level that could be delivered using renewable sources (e.g., solar, wind, and hydro) instead of burning fossil fuels.  Likewise, if we had a more autonomous fleet of cars, trucks, and buses we could provide more services on demand and there would be fewer cars parked unused for large periods of time, and consequently less resources used to build and maintain those unused vehicles. Similar arguments can be made about food productions and many other industries. Ultimately it could become practical to build systems that optimise the efficiency of the whole cycle, not just production and use but also recycle and reuse.

Are there any technologies which you think are overhyped? Why? It is common for new technologies to be initially overhyped. However, their practical use often takes much longer to develop than originally anticipated and the hype goes away. However, many end up becoming real eventually. When this happens, the speed of adoption can be surprising, especially to the people that were exposed to the initial hype. The Internet was a good example of this. It became overhyped during the “.com'' era, but now it is well established and lives up to the expectations that were set then.

I think artificial intelligence and autonomous driving are on that stage. Expectations are likely much higher than what can be delivered in the short term, but eventually will be real and practical.

Perhaps a more extreme example is smart contracts and cryptocurrencies. They are overhyped now but I expect they will find real applications and become very important in the future.

What is one unique initiative that you’ve employed over the last 12 months that you’re really proud of? One of the initiatives I have been focusing on for the last couple of years has been the technologies and product components needed to expand our platform to the cloud. The driving vision is that our users will be able to benefit from a continuous data-centric, data-sharing bus scaling all the way to the cloud that provides a level of QoS, security, and availability on-par to the ones they depend on the edge.

Our technology has typically been deployed in the “real-time edge” domain. This is where the sensor information is collected, and direct action is taken to interact with real-world electro-mechanical/physical systems. Some people have referred to this side as the “fog” domain in contraposition to the “cloud” domain.

However, there is a trend to merge these domains, more and more system span across both domains and have near real-time information exchange between them. This creates significant challenges for the architects and application developers. Our expanded technology offer will help overcome any of these challenges.

Are you leading a digital transformation? If so, does it emphasise customer experience and revenue growth or operational efficiency? If both, how do you balance the two? I focus on customer experience and improving their “operational efficiency” in terms of how efficiently they are able to develop, test and deploy complex autonomous applications using RTI’s platform.

I look at the kinds of systems they are trying to build and try to identify the software building blocks we can develop that makes them more efficient at their task. We are trying to provide users with the data-distribution plumbing they need to integrate applications no matter the platform and location of their systems. That way they can focus on their application-level problems, be it developing a medical robot, an autonomous car, or controlling airplane traffic without worrying about how to move the data around. 

What is the biggest issue that you’re helping customers with at the moment? Typically, I do not work directly with customers. I “help” them by impacting the technology we develop for them.

I interact with our customers as well as our own engineers that support them (services, support, and field engineers) to understand and anticipate the challenges they face in their applications and in their use of our technology.

The two biggest issues I am focusing on are scalability and observability.

Scalability has many dimensions, but the fundamental challenge comes from autonomous systems becoming much more complex and interconnected. Combine that with the need for availability, real-time response, deployment to more “embedded” platforms, and you end up with some very hard technical challenges. We are not working on point solutions, rather we continually analyse where limits are being hit and design ways to overcome those. Currently we are focusing on the scalability of discovery (how applications find the presence of each other) and end-to-end security (how applications are identified, permissions established, and data is authenticated and protected).

Observability are the mechanisms provided by our platform that allow users to monitor the health and performance of their overall systems and each component. It is an important aspect of keeping the availability of deployed systems. We are working on better ways to collect and distribute the monitoring information, analyse it and make it available to our users.

How do you align your technology use to meet business goals? Input and Collaboration. I seek input from all parts of the organisation to understand how customers view our technology, what are differencing aspects that help in our sales situations.  I interview customers regularly and try to understand their perspective on what where the key reasons for them to choose our technology, what other alternatives they considered, and what they think is the biggest value they are getting.

We have cross-department meetings where we perform SWOT analyses and share, sometimes debate, our perspectives. The best ideas tend to come out of this collaborative work.

Do you have any trouble matching product/service strategy with tech strategy? This is a balance that requires some effort to achieve.

The goals start with the product/service strategy, and to a large extent they drive the technology strategy.

1 2 Page 1
Page 1 of 2