Nancy Greco is a Distinguished Engineer, Technical Executive, in IBM Research in Yorktown Heights, N.Y. A Cornell graduate, with a bachelor’s degree in chemistry, she leads global teams to create AI and Edge Computing to address clients’ pressing problems in data collection, data movement, and security while getting insight and ROI from the data. She is co-leading the Consulting Robotics Services Solution development in partnership with Boston Dynamics working across IBM’s Research, Consulting, Maximo, and Cloud teams to enable roaming edge devices to do inspections in potentially dangerous environments, so humans don’t have to. She worked in IBM’s semiconductor manufacturing, as an engineer and executive, for 20 years before coming to Research.
Kay Murphy has over 25 years of business experience serving both the public and private sectors. She has delivered solutions in the aerospace, defense, manufacturing, education, general government and energy sectors. She currently leads IBM’s Global Asset Optimization Services.
Her technology background includes IoT, cognitive technology, advanced analytics, business intelligence, data warehousing, and asset and facility management. She is an expert in Maximo and the application of advanced analytics and cognitive technology to equipment maintenance.
With advancements in AI, edge computing, and robotics, there have been breakthroughs that are improving worker safety and enabling robots, like Spot, to perform tasks that human beings do not want to do in difficult or dangerous situations. Not only does this advanced technology protect humans from such jobs, but also improves efficiency and consistency in the results. Certrec was able to hold a discussion and get insights from IBM executives, Kay Murphy and Nancy Greco, who have been directly involved in the development of such technology.
Question: What are your thoughts on how IBM and Spot can help a power plant or utility today?
Kay Murphy: Talking specifically about power generation, first of all, there is the safety factor. Power generation has plenty of hazardous areas where you want to use a robot instead of a human. There is a great safety factor around that.
The second important factor mentioned by Kay was regulatory compliance.
Kay Murphy: Consider having a person doing inspections in an area using a written checklist or even a mobile device. They are checking the boxes repeatedly. As things become repetitive, humans pay less attention. It is kind of a dull job.
Kay mentioned that once an incident takes place, the regulators want to know what due diligence has been done in terms of operability of the equipment being used.
Kay Murphy: Handing the regulators hundreds or thousands of pages of checklists with all the boxes checked is not very strong evidence. However, being able to provide the visual images or thermal images, that is, all the digital information that was gathered by a robot, who does it exactly the same way every time, is extremely consistent, which is very powerful. So that is one thing.
Kay talked about equipment reliability next.
Kay Murphy: Then, there is equipment reliability, which is extremely important in power generation. If equipment goes down, you stop generating not only power, but also revenue. So, equipment reliability is super important. With Spot, we can get more information, different types of information, and different levels of fidelity in that information, which allows an organization to understand equipment health to a much greater degree to prevent crashes and to ensure the equipment is in good health.
Kay stressed on the maintenance of equipment before it stops functioning.
Kay Murphy: We all know and are very familiar with the concept that if you catch a piece of equipment before it goes down, and do some maintenance on it and take care of it, it is far less expensive and intrusive than if it goes down and you have a repair or replace issue. So, you want to avoid those replacement issues and you want to keep your equipment running as long as possible, and you do not want it to go down.
Kay used the analogy of cars to explain her point.
Kay Murphy: I always use the analogy with cars. Do you run your car to failure, or do you pay attention to it? Do you check your oil, your battery, and check everything as much as you possibly can? You want to do everything right to ensure that it does not go down, because I guarantee, when your car fails, it is at the least convenient possible time.
Nancy Greco responded to the same question, in the following way:
Nancy Greco: Yeah, and I’d like to add to that too. Many of the clients and viewers reading this will think, “but what’s really changed? Robotics has been around for a long time. Has anything changed in the past few years that makes it a very compelling case to adopt robotics?” The short answer is: absolutely! A lot has changed. Just in the past few years, you have seen the emergence of something called edge computing.
Nancy went on to explain how edge computing makes the process simpler and reduces costs.
Nancy Greco: What edge computing does is that it allows you to process the data right there on the robot, camera, or phone. Why is that so important? It reduces your cost. You don’t need a data center nearby. You don’t need to use the cloud there. So, edge computing is helping you reduce your cost and put it on very small footprints. You can put it on a drone, you can put it on the phone, or you can put it on a robot. That actually helps because that was one of the impediments when people were trying to do a lot of this “missing data” work.
Nancy explained working with semiconductors.
Nancy Greco: I worked in semiconductors and we have a lot of machines and they still go down. You will not find a more instrumented machine than a semiconductor. These machines go for a $150 million a pop. We collect tons of data. Do they ever go down? Absolutely. Why? Well, we’re missing data. Something got hot, something was making a smell, something was rattling, or something didn’t sound right.
Nancy talked about sensors that can be put on Spot to gather data.
Nancy Greco: The other big thing is that now there has been a creation of a lot of different sensors that can be put on something portable like Spot to gather this data. Thermal imaging, acoustic, we have advanced computer vision a great deal. We can not only do classification models, but we can look at a crack and tell you if it is propagating. Sensors where you’re coming in with IR. Can we smell and detect the SF6 [gas] versus having a human go out there? So, this is the second big thing, that all these sensors are coming onboard too.
The third important thing mentioned by Nancy was the changing workforce.
Nancy Greco: The third thing is that the workforce is changing. That came out with the pandemic. People are not just going to come back to work and do the dull, dirty, and dangerous anymore. They are like, “no, we’re just not going to do it.” So, what is happening is that a lot of data needs to be gathered for compliance reasons, or as Kay has said, to make sure the assets are in good health. There is no way to do it other than a human going out there with half a dozen sensors – holding the thermal camera up, holding the microphone up, and holding the gas sensor up. Quite honestly, they are:
- unlikely to do it; and
- we were finding as we started to look at it that we need consistent data.
So, when the human went out and was ten feet away, the next day they are four feet away, and then they are at an angle. We cannot create AI models with that kind of inconsistency. So, with robotics, I get consistent data, and so far, Spot has never refused to do a mission. He wasn’t tired. He might be out of energy, but he doesn’t care what weather conditions I’m putting him in. I have never had a career discussion with Spot where he says, “Really? I don’t want to read analog gauges as a career.” He just goes and does it. A bit of humor there, but there are career discussions that you have to have with employees, and going out and collecting this data is not going to get them to come and work for you.
Certrec would like to thank Kay and Nancy for their insightful perspectives.