We have a lot of equipment that’s got a really long life cycle and a slow rate of change. It’s not economical for us to go and sensorize every piece of equipment in this really vast facility. It’s probably more practical and actually quite a bit easier for us to use the mobile robotics approach.

Samantha Garrison, Deputy Director Factory Automation, GF

Manufacturing facilities are filled with thousands of inspection points, and adding fixed sensors to all these assets is not economical. With Spot dynamically bringing the sensors to their assets, the team collects valuable information about the thermal condition and analog gauge readings of pumps, motors, compressed gas systems and more. Since this equipment has a long life cycle and slow rate of change, the team can better monitor these assets over time – making better informed decisions to keep the facility running.

GF teamed up with Boston Dynamics partner Levatas to implement automated inspection missions at the facility. Building computer vision models based on gauges and motors of a variety of equipment, Levatas was able to enable Spot with the cognitive intelligence it needs to know what it’s looking at, and to make decisions based on that information. 

“With the global semiconductor shortage, it is important that operations stay moving to ensure a steady supply of product for our customer” said Samantha Garrison, GF Deputy Director of Factory Automation. With Spot on the floor collecting data, GF plans to use that information to build models that better predict planned maintenance and downtime.


Based in West Palm Beach, FL, Levatas is the leading developer of cognitive intelligence for automating industrial inspections. Levatas creates and delivers end-to-end solutions that enable robots, drones, remote sensors and camera systems to autonomously perform equipment monitoring, safety checks, and site surveillance tasks in industrial environments. Learn more at: www.levatas.com.


SAMANTHA GARRISON: We have a lot of equipment that’s got a really long life cycle and a slow rate of change. It’s not economical for us to go and sensorize every piece of equipment in this really vast facility. It’s probably more practical and actually quite a bit easier for us to implement to go and use the mobile robotics approach. 

One set of sensors as opposed to fixed sensors in every location. Here, in Burlington, this is a 200-millimeter semiconductor manufacturing facility. We make products that go into a lot of communication devices, RF technology, and the stuff that powers our connected world. 

DANIEL BRUCE: So this week, we’ve been working with a team from GlobalFoundries and Boston Dynamics and Levatas to implement automated inspection missions in GlobalFoundries’ facilities. 

SAMANTHA GARRISON: We tested Spot from Boston Dynamics. Really exciting application of mobile robotics for us in this facility. Really looking at the opportunities that mobile robotics provide for sensorization and getting more data about our facility back to our engineers and technicians. 

DAN ZUBA: Spot has an amazing athletic intelligence. It can literally go up and down stairs over all the terrains, out through snow, over different things, and get out of harm’s way. So really where Levatas comes in is Spot doesn’t know necessarily what it’s looking at. 

It knows there’s an object in front of it that it shouldn’t run into. It knows that there is something over here. But that cognitive intelligence, that’s really where Levatas comes in, is how can we train Spot to know what it’s looking at and then to make decisions based off that. Alerting somebody that this equipment is too hot or maybe it’s alerting to say, hey, there’s something trending. We’re predicting that this needs to be fixed before it goes out. 

DAVID MILONE: In the case of the area where we’re looking at here at GlobalFoundries is that they have a lot of gauges and a lot of motors. And in this case, what we could do is build models that can read those gauges and interpret that data digitally, as well as keep an eye on the thermal equipment. 

SAMANTHA GARRISON: We have a ton of peripheral equipment that is extremely important to keeping the fab running. And so from that perspective, being able to get more data about that equipment and how it’s performing is really important to our team. So Spot and the Boston Dynamics team, as well as Levatas were proving the concept that we can go and collect that data. We could navigate the robot successfully to all of those inspection points, and then use that data in the machine vision algorithms to get the information that we were looking for. 

ADRIEN PLOUFFE: That data is going to give us the opportunity to better plan maintenance time and downtime and hopefully also prevent scrap events. 

SAMANTHA GARRISON: You might have heard that there’s a semiconductor shortage, so it’s really important for us to keep operations moving and also to maintain a flow of product for our customers. The application of mobile robotics is really unique because we’re able to implement these new technologies in a way that’s flexible and also keep our facilities up and running during that implementation process. We were able to accomplish the automated mission very quickly we did that and about a half a day. And so we’re able to spend the rest of the time really understanding the capabilities that Spot has, getting to be hands-on, and thinking about additional applications that may or may not already be available or things we would like to do with Spot in the future.