W. Shane Grant

Principal Scientist (2025), Staff Scientist (2024), Senior Scientist (2022), Scientist (2018)

• Invented scalable multi-agent path planning and task assignment algorithms coordinating the movement and interaction of hundreds of robots across multiple deployments, doubling system performance.
• Led architecture through deployment of cloud/edge microservice framework to manage robot tasks and actions, including overhaul of on-robot navigation, communication, and state management, enabling 24/7 operation.
• Designed and productized customer-facing digital twin platform, providing end-to-end simulations of warehouse flow utilizing the full software stack from applications to robotic execution. Reduced quoting and design cycles from weeks to days. Empowered customers to independently model, predict, and verify operational changes without impacting production, offloading significant effort from internal teams.
• Granted 6 patents in autonomous warehouse robotics and execution systems.

Warehouse Flow Lead (2022), Robotic Management System Lead (2019)

• Built and led multidisciplinary research and engineering teams across robot management and warehouse flow, establishing a department-wide interview and recruiting framework to scale talent.
• Spearheaded kaizen continuous improvement across engineering, operations, and sales, leading to multiple process enhancements including the adoption of a monthly release cadence with phased customer rollouts.
• Designed and maintained cross‑customer performance dashboards in Grafana, analyzing operational metrics to identify bottlenecks, prioritize improvements, and guide strategic decision‑making.
• Collaborated with enterprise customers and internal product teams to gather requirements and iteratively refine simulation tooling, ensuring alignment with operational needs.
• Represented inVia Robotics as technical expert in customer calls, on-site visits, and industry events, helping close sales deals including Fortune 500 accounts.

• Completed PhD thesis on modulatory feedback in neural networks, developing a cortex-inspired framework to minimize ambiguity with a novel unsupervised `conflict learning' rule, as exemplified by its ability to model border ownership.
• Developed a state-of-the-art SLAM system for real-time, plane-based mapping using rotating LIDAR sensors for the DARPA Robotics Challenge.
• Created a point cloud library for the Neuromorphic Robotics Toolkit, a ROS-like modular framework for distributed computation.

• Computer vision and geo-spatial development, satellite image super resolution, aerial cloud simulation, and SIMD code optimization.

• Extracted and corrected imagery from raw video to match onboard sensor data. Ported legacy MATLAB code to C++.

• Implemented test fixtures in C, C++, and AppleScript for Mac Office applications and databases.

• Learned wing-to-wing decision making and workflows while shadowing leadership in Digital Services, Digital Delivery, and New Business Development.

PhD, Computer Science

BS, Computer Engineering

• Co-creator, lead developer, and maintainer of widely used modern C++ serialization library.

• Completely disassembled and rebuilt every component while refurbishing a 1974 Triumph Spitfire. Designed and fabricated custom wiring harness and daytime-running-light circuit.

• Led software systems while co-managing team, using CUDA to accelerate vision algorithms and achieve a 2nd place result in the AUVSI international competition focused on autonomous flight and target recognition.