ethz-adrl/towr

Health v2 · maintenance only

Adoption not part of the score

1079 stars · 251 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

TOWR is a lightweight, Eigen-based C++ library for trajectory optimization of legged robots. It provides reusable variables, costs, and constraints to formulate nonlinear programming problems for motions like hopping, bipedal walking, and quadruped trotting, solved via Ipopt or Snopt through ifopt.

Use cases

• generate walking or trotting trajectories for a quadruped robot
• optimize gait and step durations for a biped
• plan motion trajectories for legged robots with nonlinear programming
• visualize robot motion plans in rviz
• prototype trajectory optimization formulations for legged locomotion
• use trajectory optimization for physics-based animation

When to choose

• you need fast (<100ms) trajectory optimization for monoped, biped, or quadruped motions
• you want an extensible, lightweight C++ base-set of variables, costs, and constraints for legged locomotion
• you work in a ROS/catkin environment and want rviz visualization via xpp
• you want to use high-performance NLP solvers like Ipopt or Snopt through ifopt

When to avoid

• you need trajectory optimization for wheeled, aerial, or manipulator robots rather than legged systems
• you need a actively developed project with frequent updates (last release 2023, largely research code)
• you want a pure-Python or non-ROS workflow without building from source
• you need a full robot controller, not just motion plan generation

Facets

library · maturity maintenance

simulation math sdk robotics cpp trajectory-optimization legged-robots motion-planning nonlinear-programming ipopt snopt ifopt eigen ros locomotion gait-optimization algorithms research linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="ethz-adrl/towr")