Max Robotics — Global Service Robot Directory + US Certification

About

T. Oshima is a leading researcher in robotics and hardware acceleration, specializing in real-time object pose estimation and motion planning for industrial robots. Their most significant contributions center on dramatically improving picking robot throughput by developing FPGA-based accelerators for the computationally intensive iterative closest point (ICP) algorithm. Oshima’s 2020 work on an SoC-FPGA-based ICP accelerator (28 citations) directly addresses the bottleneck of low picking speed in conventional systems. They further advanced the field with a hierarchical graph-reusing k-NN search technique (15 citations) and a 4.8× faster FPGA-based ICP accelerator (11 citations), validated using Amazon Picking Challenge datasets. Oshima also pioneered motion planning for dual-arm robots, creating an FPGA accelerator capable of handling graphs with over 10 million edges (10 citations). Their work extends to human-robot collaboration, including a novel RGB/infrared camera fusion method for multi-person pose estimation in low-light environments (10 citations). By combining algorithmic innovation with custom hardware design, Oshima has enabled faster, more reliable robotic manipulation, with direct applications in warehouse automation and collaborative robotics.