How Locus is getting a grasp on one of robotics biggest challenges: manipulation

Locus Robotics has overcome many robotic challenges in real-world environments. Its autonomous mobile robots (AMRs) have navigated countless miles and aided in over 6 billion picks. Now it is targeting manipulation.

The company entered robotic picking in March 2025 with the unveiling of Array, a mobile manipulator with a custom vision-guided picking arm. Over a year later, Locus acquired Nexera Robotics, known for its NeuraGrasp gripping technology.

Roy Belak, senior vice president of robotic grasping at Locus and former CEO of Nexera, said that Nexera had spent years refining its soft end effectors before meeting Locus, going through at least six product generations. He told The Robot Report that they encountered Locus at MODEX at a fortuitous moment, and the history "kind of writes itself from there."

The timing was ideal for both. Locus wanted to handle more SKUs, and Vancouver-based Nexera was ready to scale. "We want to see this technology deployed at scale," Belak said, noting he was excited as an innovator to see it widely adopted.

What does NeuraGrasp bring to Locus Array?

Nexera met Locus in 2025 when Locus was planning Array's early market release. At that time, Array had a standard suction-cup gripper. Belak said the companies spent significant time integrating Nexera's technology into Array's development.

Fortunately, Array was already well prepared, having air and power supplies that Nexera's technology required. "There was a lot of work done prior to the acquisition to make sure the technology could be slipped into the development stream," Belak recalled. He added that missing that window by a year would have made changes very difficult.

Belak did not give an exact timeline for customers accessing Nexera's technology, but he is optimistic. He expects that the next public display of Array will likely feature Nexera technology.

Currently, Locus is in early deployment of Array with its original gripper, so Nexera hasn't yet seen the data influx. "There's a long line of customers with their hands up that are really excited about Array," Belak said, noting it is somewhat intimidating because the integrated product will go to many places, and then the data will come back for optimization.

Suction versus pinch picking

Regarding robotic manipulation, Belak said much of it falls on a spectrum between suction-based and pinch grasping. "Suction is very good, particularly in warehouse operations," he said. "Sixty to 70 percent of the time, it's all you need. It's when you get to that 30 to 40 percent with high coverage ratios."

NeuraGrasp's soft gripping technology bridges this gap by combining suction and pinching. "As an industry trend, it's 100% recognized that pinch grasping is likely going to be very influential in the future of robotic grasping," Belak said. He added that suction gets you a good portion of the way, and "if it ain't broke, you don't throw it away."

Locus focuses on reliability and quality picks, not speed.

"In the industrial environment, reliability is everything," Belak said. Suction works well but has limitations; pinch grasping can vary in reliability. It's important to make robots that simply succeed at tasks. The top priority is reliability, because breakdowns or missed picks negate performance. Quality is essential too.

"There's a big differentiation between being able to do something, and being able to do the same tasks with high quality," Belak said. Locus Origin bots are known for quality, and they want to bring that ethos to picking. Historically, picking claims focus on ROI without understanding quality. "We can do many picks per hour, but what people don't focus on is how good are those picks," he continued.

Locus aims to avoid product damage, double picks, and misspicks. These aren't often advertised but can quickly eat into customer costs. "Once they have the quality down, then they can worry about increasing speed," Belak said.

What's holding back robotic picking?

Belak identified two roadblocks: AI development and giving robots a sense of touch. The industry is good at building hands and electromechanical structures, but pinch grasping intelligence lags behind. "[The challenge] is 100% the proprioception and the data metrics and the feedback mechanisms," he said.

The robotics industry has raised millions for physical AI, but human brains evolved over billions of years. Generating data helps, but it highlights the lack of tactile sensing. "There is one piece of the physical system which is missing, which is the sense of touch on the robots, and particularly robust, reliable touch that doesn't degrade after very limited usage," Belak said. He noted simulated data won't help much with this problem.

Nexera currently uses a sense of touch and plans to advance it. While Locus is early with Array, its manipulation roadmap extends beyond. "Robotic manipulation in general—whether it's Array, other versions, or picking benches—we have identified that as a pillar of innovation," Belak said.

Editor's note: Rick Faulk, CEO of Locus Robotics, will speak at RoboBusiness 2026 in Santa Clara, Calif., on Oct. 20-21, part of the panel "The RaaS Playbook: Pricing, Service & Scale."