OSARO is building SightWorks grasp-depth sensing into fulfillment picking

OSARO secured USPTO patent 12,236,340 for computer-automated robot grasp depth estimation on February 25, 2025. The patent covers a method for estimating object depth from images and robot arm poses gathered during successful and failed grasp attempts, giving OSARO a specific technical claim inside warehouse piece picking.

The San Francisco company connects the patent to SightWorks, its vision software for fulfillment picking, kitting, bagging, depalletizing, and induction. OSARO describes the method as a way to reduce dependence on specialized depth sensors, which can be material in fulfillment cells that face reflective objects, plush toys, bagged apparel, and other categories that frustrate generic grasping.

SightWorks suggests pick points and item orientation for piece picking and bagging. For mixed-case depalletizing, it recognizes cases, detects damaged cartons, and flags pallet protrusions. OSARO describes the product as software installed on an industrial computer with an NVIDIA GPU, with robotic motion control translating perception into picks and placements.

AutoModel is the SKU update channel. It updates models for new SKUs without manual labeling, targeting packaging changes and end-effector changes. OSARO made SightWorks available to integrators and 3PLs in June 2023, positioning the software across depalletization, piece picking with AMR or ASRS systems, kitting, and induction work.

The competitive field includes Plus One Robotics, RightHand Robotics, Covariant/Amazon, Mujin, Berkshire Grey-style fulfillment systems, Pickle Robot, and integrators using conventional 3D vision. OSARO's distinction is a software-first grasping stack that can be sold through integrators and adapted to hard SKU categories without requiring every customer to build a new perception system.

Public material does not show production pick accuracy by SKU class, customer-level intervention frequency, per-cell throughput by workflow, tool-change reliability, deployed cell count by integrator, customer renewal terms, or post-install support response time. The strategic test is whether OSARO can turn grasp-depth learning into lower exception rates in live fulfillment cells. If SightWorks improves difficult-item handling without specialized sensing cost, OSARO becomes a fulfillment perception layer rather than a one-cell picking vendor.

Have a robotics update Korthos should review? Send news, deployments, product releases, funding rounds, research, or media to tips@korthos.xyz or reach out on X at @agkorthos.