Atlas set for chip factories as GlobalFoundries scales Spot
GlobalFoundries will test Boston Dynamics’ Atlas in early 2027 as the first non-automotive customer, while two Spot robots in Singapore complete 90,000-plus inspections at a 99% success rate.

Why it matters
- Atlas on-site testing at GlobalFoundries is scheduled to begin in early 2027.
- GlobalFoundries is the first customer outside the automotive industry to test Atlas.
- Two Spot robots in Singapore completed more than 90,000 inspections since April, covering nearly 900 kilometers.
- The Singapore deployment achieved a 99 percent mission success rate.
- GlobalFoundries plans to expand Spot across four sites and more than double its inspection workload.
GlobalFoundries will begin on-site testing of Boston Dynamics’ humanoid robot Atlas in early 2027, becoming the first customer outside automotive manufacturing to try the machine. It is also the first semiconductor manufacturer to join Boston Dynamics’ humanoid robot early adopter program.
Boston Dynamics said Tuesday, local time, that it had expanded its partnership with the U.S. chipmaker. Boston Dynamics, the robotics subsidiary of Hyundai Motor Group, will test Atlas and deploy more four-legged Spot robots across GlobalFoundries facilities worldwide.
The arrangement links two stages of factory robotics. Spot already performs inspection work that GlobalFoundries can scale, while Atlas still faces task selection, integration work and site-specific training. It creates a working baseline against which the chipmaker can judge a humanoid system.
What changes in early 2027?
GlobalFoundries and Boston Dynamics will identify manufacturing tasks suited to humanoid robots during the planned on-site test. They will also assess integration requirements, then train Atlas for the specific operations they select.
The announcement does not name a factory task or promise an immediate production rollout. Its sequence links the work: define the operation, work out what the robot must connect with, and train the machine for that setting. For outside observers, those details will determine whether the trial addresses a repeatable process or remains a broad demonstration.
For now, early 2027 is a test date, not an Atlas deployment date.
Why is Spot expanding first?
GlobalFoundries plans to expand Spot deployments across four facilities:
- Malta, New York
- Burlington, Vermont
- Dresden, Germany
- Singapore
The chipmaker has used Spot at its Vermont facility for more than four years. It recently extended the four-legged robot to Singapore, so the expansion builds on a long-running installation rather than introducing Spot to a new user.
At the Singapore facilities, two Spot robots have traveled nearly 900 kilometers and completed more than 90,000 inspections in sub-factory areas since April. GlobalFoundries reports a 99 percent mission success rate for those missions.
GlobalFoundries also plans to more than double its inspection workload using Spot’s advanced visual inspection capabilities. That statement refers to workload, not necessarily a doubling of the robot fleet; the announcement gives no robot-count target for the expansion. The operational evidence remains concrete, though: one deployment has covered more than 90,000 inspections across nearly 900 kilometers at a 99 percent success rate.
Spot supplies the measurable near-term expansion, while Atlas remains scheduled for a separate evaluation in early 2027.
How does inspection data support maintenance?
Pradip Singh, chief manufacturing officer at GlobalFoundries, said:
“Spot has helped us shift from reactive to predictive maintenance, collecting data across thousands of inspection points to prevent downtime and keep our fabs running more efficiently.”
“Scaling Spot and testing Atlas are the next steps in our journey to AI-enabled manufacturing, putting physical AI to work across our operations worldwide.”
Singh described a change from reactive to predictive maintenance. Spot collects data across thousands of inspection points, and GlobalFoundries uses that record to anticipate problems, prevent downtime and keep its fabs running more efficiently.
The described mechanism gives robots a production role built around repeated checks and accumulated maintenance data. The announcement does not specify which assets Spot inspects, what defects it detects or how workers act on each result.
The 99 percent figure measures mission success, not the accuracy of every finding. It also does not show false-positive or false-negative rates. Those omissions leave open how much human review remains inside the shift from reactive to predictive maintenance.
The current proof covers completed robot missions, not a complete measure of inspection quality.
What must Atlas prove?
Atlas enters a customer relationship that already has operating data from Spot. GlobalFoundries will test whether a humanoid can take on suitable manufacturing tasks, integrate with the factory environment and train for specific operations.
Zack Jackowski, chief product and technology officer at Boston Dynamics, said:
“GlobalFoundries was one of our earliest customers to adopt Spot, and we are excited to see the program continue scaling across the globe.”
“Innovative companies like GF are realizing real operational value today while collecting the kind of data that will be extremely useful toward future manufacturing innovation in the next few years.”
Jackowski tied those statements to present operational value and future manufacturing innovation. His first quote describes GlobalFoundries as an early Spot customer and frames the partnership’s next phase as global scaling. His second says deployments are already generating data that should be useful over the next few years.
Neither executive said which Atlas task GlobalFoundries selected. The announcement also provides no details on robot configuration, integration standards, safety criteria, commercial terms or the number of humanoids involved. Boston Dynamics is presenting Atlas through a customer trial rather than a completed factory rollout.
Boston Dynamics is testing a route from proven inspection work toward broader humanoid operations.
What do the partnership terms reveal?
GlobalFoundries holds two distinct positions in Boston Dynamics’ robotics strategy. It is the first semiconductor manufacturer and the first non-automotive customer in the humanoid early adopter program, while also operating Spot for more than four years in Vermont.
The Spot footprint gives the partnership a current use case with measured output. The four planned locations — Malta, Burlington, Dresden and Singapore — extend that work across GlobalFoundries’ global facilities. Atlas supplies the research question: whether the humanoid program can produce useful tasks in a new industrial setting.
The companies frame physical AI as a working strategy, not a distant concept. Singh called scaling Spot and testing Atlas the next steps toward AI-enabled manufacturing across GlobalFoundries’ operations. Jackowski said customer deployments generate present-day operational value and data for future manufacturing innovation.
The arrangement has operating evidence now, but its commercial scale for Atlas remains undefined.
What should readers watch next?
The first checkpoint comes in early 2027, when the partners begin on-site Atlas testing. The useful evidence will be the selected manufacturing task, the integration requirements they uncover and whether training produces repeatable operations.
For Spot, readers should separate plans from completed work. The expansion covers Malta, Burlington, Dresden and Singapore, and the chipmaker intends to more than double its inspection workload. The 99 percent success rate and 90,000-plus inspections apply to the existing Singapore deployment since April, not to every expanded site.
The announcement leaves several commercial questions open. It does not disclose prices, fleet commitments, Atlas purchase plans or a timetable beyond the start of testing. It also does not specify how GlobalFoundries will compare Atlas with Spot or decide which tasks warrant humanoid hardware.
The Korea Times said the original report used generative-AI assistance and was edited by its staff.
If Atlas completes repeatable factory work at GlobalFoundries, Boston Dynamics will have evidence that its humanoid program extends beyond automotive manufacturing. Until then, Spot’s inspection record is the proven result; the larger test waits for early 2027.
Source: feed.koreatimes.co.kr
More from Sophie Lindqvist
Show full bio
Staff writer covering regulation, policy and their impact on artificial intelligence and machine learning at AI In Context.
749 articles
Related articles
- Boston Dynamics Swapped Atlas's Feet for Scooter Wheels
- Boston Dynamics' Atlas lands 360-degree spinning jump on wheels
- Apollo 2 Humanoids Sort Mercedes Parts at 96% in DeepMind Test
- Industrial Next promises factory retooling in minutes, with BMW as first test
- Agility Robotics Debuts on Nasdaq as First US Pure-Play Humanoid Stock