Deep Learning for Object Recognition & Pose Estimation

The contract focuses on the development and training of deep learning models to perform detection, classification, and six-degree-of-freedom pose estimation of laboratory equipment using visual data. The work requires advanced computer vision techniques to analyze imaging inputs and accurately identify and locate scientific instruments in complex lab environments, with precision critical to automation and operational efficiency. The project is issued as a subcontract under the NAICS code 511210, indicating its classification within software publishing and custom software development, and is managed by Brookhaven National Laboratory under the U.S. Department of Energy. The solicitation was posted on June 26, 2026, with a firm deadline for responses on July 13, 2026, at 9:00 PM Eastern Time. Performance of the work is to be carried out in Upton, New York, with a ZIP code of 11973, aligning with the physical location of the facility hosting the laboratory systems and vision data collection. The contract does not specify a set-aside type or point of contact, implying the opportunity is open to qualified entities without restrictions, and interested parties must submit proposals through the official SAM.gov portal linked in the posting.