Antoine Bambade

About

Antoine Bambade is a leading researcher at the intersection of numerical optimization and robotics, whose work is redefining how robots plan, control, and learn in real time. His primary contributions lie in developing high-performance solvers for quadratic programming (QP) and trajectory optimization, with a focus on real-world robotic applications. Bambade is the principal architect behind the PROX-QP and ProxQP solvers, which have become essential tools for whole-body control, motion planning, and estimation—cited over 60 times collectively for their speed and reliability. He also pioneered constrained Differential Dynamic Programming (DDP) using primal-dual augmented Lagrangian methods, enabling efficient handling of complex constraints in nonlinear optimal control. More recently, Bambade has advanced differentiable collision detection through randomized smoothing, allowing gradient-based learning to incorporate geometric constraints directly. His work on QPLayer further bridges optimization and deep learning by enabling efficient differentiation of convex QP layers. With over 135 citations across his most influential papers, Bambade’s contributions are shaping the next generation of real-time, constraint-aware robotic systems.

Research Focus

Key Achievements

Top Papers

1. 1
   PROX-QP: Yet another Quadratic Programming Solver for Robotics and beyond

   54 citations · 2022
2. 2
   Constrained Differential Dynamic Programming: A primal-dual augmented Lagrangian approach

   32 citations · 2022
3. 3
   Differentiable Collision Detection: a Randomized Smoothing Approach

   26 citations · 2023
4. 4
   ProxDDP: Proximal Constrained Trajectory Optimization

   12 citations · 2025
5. 5
   ProxQP: an Efficient and Versatile Quadratic Programming Solver for Real-Time Robotics Applications and Beyond

   9 citations · 2025
6. 6
   QPLayer: efficient differentiation of convex quadratic optimization

   2 citations · 2023

Key Collaborators

Contact & Links