1 month ago
Berkeley, CA, USAIntern
Responsibilities
- Develop and tune PID, MPC, optimal, and robust controllers for dynamic, nonlinear systems.
- Build and validate physical models for simulation, hardware-in-the-loop testing, and autonomy.
- Train and deploy supervised or reinforcement learning models for perception, planning, or adaptive control.
- Integrate algorithms with firmware and cloud teams while ensuring real-time safety and stability.
- Profile, optimize, and verify performance under latency, jitter, and compute constraints.
Requirements
- Working toward a bachelor’s degree or higher in Computer Science, Robotics, or a related field.
- U.S. permanent residency is required.
- Preferred experience includes hobby or personal projects, maker work, LiDAR integration, Python, C++, PyTorch, JAX, MPC, OSQP, CasADi, EKF, UKF, factor graphs, system identification, PPO, SAC, safety shields, ROS2, ONNX, TensorRT, latency and jitter profiling, PCB design, and embedded systems.
- Ability to work at the pace and intensity of an early-stage startup.
Benefits
- $25-$35 per hour plus relocation assistance and company perks.
- Spring internship from January through May, depending on the student’s school schedule.
About Terranova
Terranova builds car-sized terraforming robots to lift land out of flood zones. These robots inject a wood slurry deep underground, precisely reshaping property and creating new land, without surface disturbance. This elevation is permanent and provides proactive protection for all types of infrastructure, wetlands, and new development projects. Our three robot and one mothership system can lift an acre by a foot every day. Backed by Outlander VC, Congruent Ventures, Galvanize’s seed fund, GoAhead Ventures, and angels like Palmer Luckey. Please contact us at contact@terranova.inc for a quote on your elevation project.