Warehouse robots can reach the shelf. The harder test is grabbing messy goods
Locus Array at a fulfillment tote, NeuraGrasp pressing its soft membrane around mixed awkward items instead of using a simple suction cup.📷 AI-generated image / TECH&SPACE
- ★Locus is acquiring Nexera Robotics so NeuraGrasp can replace Array's suction gripper.
- ★The gripper combines a soft membrane, sensing, computer vision and AI for a wider SKU range.
- ★Its value will be measured in fewer exceptions and failed picks, not in a polished demo video.
Warehouse robots rarely fail because they cannot drive down an aisle. They fail because the object waiting at the end of that aisle is wrapped in plastic, half-crushed, glossy, flexible, tilted, or simply not where the system expected it to be. That is why Locus Robotics' acquisition of Nexera Robotics is more than a feature upgrade. It points at the awkward part of automation: real contact with real goods.
According to The Robot Report, Nexera's NeuraGrasp is expected to replace the suction-cup gripper on Locus Array and expand the range of e-commerce SKUs the system can pick. Locus Robotics has already built its business around autonomous warehouse robots, but Array moves into the harder layer of the job: mobile manipulation, where the robot must not only reach a shelf or tote, but decide how to grip the specific item in front of it.
A suction cup makes sense where surfaces are flat, smooth and predictable. It is fast, mechanically simple and useful for a certain class of packages. Fulfillment, however, is not a catalog of polite boxes. It includes soft bags, dented cartons, labels that catch, tiny products, cosmetics, apparel, lightweight goods and items pressed together at bad angles. In that environment, grip force alone is not the point. Adaptive contact is.
NeuraGrasp replaces the suction cup and targets the awkward goods warehouses cannot neatly stage for robots.
Close forensic view of the gripper contact: soft membrane conforming around a dented carton and plastic bag while vision overlays map candidate grasp points.📷 AI-generated image / TECH&SPACE
NeuraGrasp combines a soft membrane, onboard sensing, computer vision and AI. In practical terms, the system is trying to see, feel and judge the pick before the item is lifted. The source report says Array handles SKUs up to 5 lb, or about 2.2 kg, placing it in the light and mid-range e-commerce goods zone rather than the pallet-scale industrial handling lane.
The claims still need a cold reading. The Robot Report coverage points to a broader SKU range, but it does not provide speed, accuracy, uptime or category-level compatibility numbers. That makes the direction important, not yet proven at scale. A controlled demonstration, with clean lighting and cooperative object placement, is not the same thing as a busy warehouse with reflections, dust, humans nearby and bins packed by imperfect processes.
The business signal is sharper. Locus did not merely license a tool; it acquired Nexera Robotics. That suggests it sees manipulation as core product capability, not a peripheral accessory. Array has already won a 2026 RBR50 Robotics Innovation Award, but awards do not move orders. The useful version of this technology will be the one that knows when to try, when to change grip strategy and when to hand the exception back to a person.
So this is less a story about one gripper and more a marker for where warehouse automation is going. Driving to the bin is becoming table stakes. The expensive question is whether the hand at the end of the robot can survive the messy physics of packaging often enough to make the whole trip worthwhile. That is where NeuraGrasp, Locus Array and the broader promise of mobile manipulation will be judged.
