Portrait of Roland Schwan

Roland Schwan

Postdoctoral Researcher

Dartmouth College, Computer Science

roland.schwan@dartmouth.edu

CV

About

I'm a postdoctoral researcher in the Department of Computer Science at Dartmouth, hosted by Brian Plancher. I completed my PhD studies in 2026 at the Laboratoire d'Automatique (Predictive Control Lab) and the Risk Analytics and Optimization Chair (RAO) under the supervision of Colin N. Jones and Daniel Kuhn. was a member of and funded by NCCR Automation. I received the B.S. degree in electrical engineering and information technology from ETH Zurich in 2019, and the M.S. degree in control systems from Imperial College London in 2020. During summer 2024, I visited Paul Goulart's lab at the University of Oxford.

Research

I am interested in high-performance numerical optimization and its applications to robotics and related fields, especially model predictive control, multistage optimization, and quadratic programming. More recently, my work has focused on exploiting problem structure for GPU-accelerated algorithms and high-quality numerical implementations. In prior work, I also worked on the verification of neural network-based controllers.

Software

PIQP

A proximal interior-point solver for dense and sparse quadratic programs, designed for embedded and real-time applications. PIQP supports fast repeated solves, avoids dynamic memory allocations during re-solves, and provides interfaces to Python, Matlab/Octave, and R.

Code Solver paper Multistage paper Parallel KKT paper

socu

A Python toolbox for GPU-accelerated solution of structured sparse linear systems arising in model predictive control. It uses custom CUDA kernels to exploit block-tridiagonal structure in long-horizon problems.

Code Paper

EVANQP

A verification framework for approximate neural networks and parametric quadratic programs. EVANQP compares learned policies against optimization-based controllers, searches for worst-case approximation errors, and can be used to certify stability over a given domain.

Code Paper

Publications

Preprints

  1. GPU-Accelerated Cholesky Factorization of Block Tridiagonal Matrices

    R. Schwan, D. Kuhn, and C. N. Jones

    Preprint, 2026

    Paper Code

  2. Parallel KKT Solver in PIQP for Multistage Optimization

    F. Song, R. Schwan, Y. Chen, and C. N. Jones

    Preprint, 2025

    Paper Code

Journal Papers

  1. Distributed Real-Time Cooperative Model Predictive Control

    Y. Jiang, K. Fedorová, R. Schwan, J. Oravec, and C. N. Jones

    IEEE Transactions on Automatic Control, 2026

    Paper

  2. On continuation and convex Lyapunov functions

    W. Jongeneel and R. Schwan

    IEEE Transactions on Automatic Control, vol. 69, no. 10, pp. 6895-6906, 2024

    Paper

  3. Stability Verification of Neural Network Controllers Using Mixed-Integer Programming

    R. Schwan, C. N. Jones, and D. Kuhn

    IEEE Transactions on Automatic Control, vol. 68, no. 12, pp. 7514-7529, 2023

    Paper Code

Conference Papers

  1. LaOPT: A Native C++ Optimal Control Toolbox for High-Performance Implementations, and Application to Racing

    J. Waibel*, R. Schwan*, and C. N. Jones

    IEEE Conference on Control Technology and Applications (CCTA), 2026

  2. Exploiting Multistage Optimization Structure in Proximal Solvers

    R. Schwan, D. Kuhn, and C. N. Jones

    IEEE Conference on Decision and Control (CDC), 2025, pp. 4677-4683

    Paper Code

  3. Cooperative Distributed Model Predictive Control for Embedded Systems: Experiments with Hovercraft Formations

    G. Stomberg*, R. Schwan*, A. Grillo, C. N. Jones, and T. Faulwasser

    International Conference on Robotics and Automation (ICRA), 2025, pp. 11377-11383

    Paper

  4. On identifying the non-linear dynamics of a hovercraft using an end-to-end deep learning approach

    R. Schwan, N. Schmid, E. Chassaing, K. Samaha, and C. Jones

    IFAC Symposium on System Identification, vol. 58, 2024, pp. 289-294

    Paper

  5. PIQP: A Proximal Interior-Point Quadratic Programming Solver

    R. Schwan, Y. Jiang, D. Kuhn, and C. N. Jones

    IEEE Conference on Decision and Control (CDC), 2023, pp. 1088-1093

    Paper Code

  6. Physics-Informed Machine Learning for Modeling and Control of Dynamical Systems

    T. X. Nghiem, J. Drgona, C. N. Jones, Z. Nagy, R. Schwan, B. Dey, A. Chakrabarty, S. Di Cairano, J. A. Paulson, A. Carron, M. N. Zeilinger, W. Shaw Cortez, and D. L. Vrabie

    American Control Conference (ACC), 2023, pp. 3735-3750

    Paper

  7. Optimal Thrust Vector Control of an Electric Small-Scale Rocket Prototype

    R. Linsen, P. Listov, A. de Laiarte, R. Schwan, and C. N. Jones

    International Conference on Robotics and Automation (ICRA), 2022, pp. 1996-2002

    Paper

Dissertations

  1. Numerical Methods for Optimization and Control: From Verification to GPU-Acceleration

    R. Schwan

    PhD thesis, EPFL, 2026

    Paper

  2. Data-Driven Economic Model Predictive Control

    R. Schwan

    MSc thesis, Imperial College London, 2020

*Equal contribution

Teaching