𝐶𝑎𝑟𝑎𝑚𝑖𝑧𝑎𝑟𝑢'𝑠 𝐵𝑙𝑜𝑔 lists

Automatic Differentiation

anonymous@undisclosed.example.com (Anonymous) — Fri, 24 Nov 2023 16:39:39 +0000

Automatic Differentiation

Automatic Differentiation implementation

GitHub Implementation / Slides

→ Read more...

2026/06/07 21:11 · Horea Caramizaru · 0 Comments · 0 Linkbacks

automatic-differentiation, algorithmic-differentiation, tu-wien, 2020

TinyAD: Automatic Differentiation in Geometry Processing Made Simple

→ Read more...

2024/03/24 22:27 · Horea Caramizaru · 0 Comments · 0 Linkbacks

patrick-schmidt, janis-born, david-bommes, marcel-campen, leif-kobbelt, geometric-processing, automatic-differentiation, algorithmic-differentiation, tinyad, ad, 2022

Matrix Calculus for Machine Learning and Beyond

Course Description

We all know that calculus courses such as 18.01 Single Variable Calculus and 18.02 Multivariable Calculus cover univariate and vector calculus, respectively. Modern applications such as machine learning and large-scale optimization require the next big step, “matrix calculus” and calculus on arbitrary vector spaces.

This class covers a coherent approach to matrix calculus showing techniques that allow you to think of a matrix holistically (not just as an array of scalars), generalize and compute derivatives of important matrix factorizations and many other complicated-looking operations, and understand how differentiation formulas must be reimagined in large-scale computing. We will discuss reverse/adjoint/backpropagation differentiation, custom vector-Jacobian products, and how modern automatic differentiation is more computer science than calculus (it is neither symbolic formulas nor finite differences).

→ Read more...

2024/03/24 15:19 · Horea Caramizaru · 0 Comments · 0 Linkbacks

math, optimization, derivatives, automatic-differentiation, algorithmic-differentiation, ml, alan-edelman, steven-johnson, mit, mit-opencourseware, 2023, course

The Elements of Differentiable Programming

Abstract

Artificial intelligence has recently experienced remarkable advances, fueled by large models, vast datasets, accelerated hardware, and, last but not least, the transformative power of differentiable programming. This new programming paradigm enables end-to-end differentiation of complex computer programs (including those with control flows and data structures), making gradient-based optimization of program parameters possible. As an emerging paradigm, differentiable programming builds upon several areas of computer science and applied mathematics, including automatic differentiation, graphical models, optimization and statistics. This book presents a comprehensive review of the fundamental concepts useful for differentiable programming. We adopt two main perspectives, that of optimization and that of probability, with clear analogies between the two. Differentiable programming is not merely the differentiation of programs, but also the thoughtful design of programs intended for differentiation. By making programs differentiable, we inherently introduce probability distributions over their execution, providing a means to quantify the uncertainty associated with program outputs.

Arxiv Link

→ Read more...

2024/03/24 14:35 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mathieu-blondel, vincent-roulet, optimization, derivatives, automatic-differentiation, algorithmic-differentiation, differentiable-programming, ml

So You Think You Know How to Take Derivatives? | Steven Johnson

→ Read more...

2023/11/24 20:14 · Horea Caramizaru · 0 Comments · 0 Linkbacks

optimization, derivatives, automatic-differentiation, algorithmic-differentiation

Introduction to Algorithmic Differentiation (AD I)

Course AD

→ Read more...

2023/11/02 12:02 · Horea Caramizaru · 0 Comments · 0 Linkbacks

course, math, ad, algorithmic-differentiation, automatic-differentiation, video-course

The Art of Differentiating Computer Programs

→ Read more...

2023/11/02 11:59 · Horea Caramizaru · 0 Comments · 0 Linkbacks

ad, algorithmic-differentiation, automatic-differentiation, book

What is Automatic Differentiation?

→ Read more...

2023/10/16 18:40 · Horea Caramizaru · 0 Comments · 0 Linkbacks

automatic-differentiation, algorithmic-differentiation

Journal

anonymous@undisclosed.example.com (Anonymous) — Sun, 31 Dec 2023 14:46:51 +0000

Journal

Reconnecting with the Younger Me

→ Read more...

2025/09/30 14:01 · 0 Comments · 0 Linkbacks

journal, 2025

Letting Go...

In the end,
the whole of life is a continuous act of letting go…
but this time,
I got the chance to say goodbye.

→ Read more...

2025/07/07 18:44 · Horea Caramizaru · 0 Comments · 0 Linkbacks

journal, edinburgh, old-college, the-university-of-edinburgh

My first watercolor portrait...

→ Read more...

2025/05/04 02:32 · Horea Caramizaru · 0 Comments · 0 Linkbacks

watercolor, paint, journal, 2021

New objective :)

I've got 3.5 years to master the piano like a pro. 😄 It's never going to happen, but I can still try.

Edinburgh Uni practice room

→ Read more...

2025/04/13 13:21 · Horea Caramizaru · 0 Comments · 0 Linkbacks

journal, song, 2025

Silence...

Silence… there is so much silence around me tonight, and nobody is here to feel it, to scent it, to hear it.

Apart from you, stranger…

→ Read more...

2025/04/10 23:25 · Horea Caramizaru · 0 Comments · 0 Linkbacks

edinburgh, journal, 2025

Imprinted

→ Read more...

2025/02/17 23:56 · Horea Caramizaru · 0 Comments · 0 Linkbacks

journal, poem, 2025

Gil Strang's Final 18.06 Linear Algebra Lecture

After being with MIT for almost 66 years, at the age of 88, the legendary Prof. Strang had today his last Linear Algebra Lecture.

Thank you for the great service you have done to the math community and thank you, for making me and many other math enthusiasts, enjoy so much this topic! To this day, I remember vividly the time when I found your lectures. It was ~2006 and MIT OpenCourseWare was still in its infancy… I remember being stuck in front of my computer for hours enjoying your lectures! My life has never been the same since then

→ Read more...

2023/05/15 12:00 · 0 Comments · 0 Linkbacks

journal, strang, mit, mit-opencourseware, 2023

Eu nu strivesc corola de minuni a lumii

(I do not crush the crown of this world’s wonders)

→ Read more...

2023/03/18 00:08 · Horea Caramizaru · 1 Comment · 0 Linkbacks

journal, poem

Optimization

anonymous@undisclosed.example.com (Anonymous) — Sun, 24 Dec 2023 14:55:02 +0000

Optimization

GIAN course on Advances in Mixed-Integer Nonlinear Optimization

→ Read more...

2025/12/15 13:20 · Horea Caramizaru · 0 Comments · 0 Linkbacks

optimization, mix-integer-optimization, nonlinear-optimization, sven-leyffer, pietro-belotti, ashutosh-mahajan, 2024

Optimization and Learning for Robot Control 2025

→ Read more...

2025/11/17 22:38 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, lqr, ilqr, task-space-control, ddp, ml, optimal-control, mpc, optimization, qp, andrea-del-prete, 2025

Memmo Winter School 2019

Here is a list of topics that will be covered during this school

Movement primitives
Pinocchio: kinematics and derivatives in C++ and python
Optimal control solvers
Contact planner
Walking robot navigation and state estimation
Motion planning for legged locomotion (tentative title, probably focus on quadrupedal locomotion)
Motion planning workflow for the atalante exoskeleton
TALOS control system architecture and whole body controller
Deep RL
QP-based task-space inverse
Dynamics: theory and implementation
Robust optimization with application to task-space inverse dynamics

→ Read more...

2025/11/17 17:41 · Horea Caramizaru · 0 Comments · 0 Linkbacks

Real-time constrained trajectory optimization in robotics: theory, implementation, and application

Abstract

Robotics has been a momentous field of research and development over the past sixty years, since the computer chip revolution in the 1960s. Realising the promise of the field entails research into motion planning and control of robotic systems in many scenarios: from fixedbase robots (e.g. industrial arms) to floating-base wheeled orlegged robots. For such schemes to be able to deal with more complex robots, they must plan ahead and satisfy physical and operational constraints, and do so, hopefully, in a reactive manner. For many years, they were rather simple, limited to simple robots doing simple things. Greater complexity and capability was unlocked with more sophisticated planners and controllers, that had access to better models of their subservient systems (geometries, inertia, the existence of contact). An adequate framework to reconcile the domain's requirements is that of optimal control. It allows for predictive models of robot behaviour over given time horizons! , and constraint satisfaction, while optimising for a given performance metric. As most optimal control problems cannot be solved in closed-form, we resort to numerical methods. This numerical optimal control has a proven track record for online motion generation and control on legged robots with real-time requirements (albeit mostly while using simplified models of the robot and environment). However, it typically leads to large-scale mathematical optimisation problems with thousands of variables – a computationally expensive endeavour. Thus, its use in robotics has relied on two axes of progress: faster chips, and efficient, structure-exploiting algorithms (with carefully engineered implementations). This thesis focuses on the latter axis: development of more performant, real-time capable solvers for numerical optimal control, with the objective of “on-the-fly” complex motion generation the for predictive control of sophisticated robots.

→ Read more...

2025/06/09 10:53 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, optimal-control, optimization, mpc, wilson-jallet, inria, cnrs, phd, 2024

Winter School on Numerical Methods for Optimal Control of Nonsmooth Systems

Monday, February 03, 2025, 8:00 - Wednesday, February 05, 2025, 16:00

Mines Paris-PSL, 60 Boulevard Saint-Michel, 75006 Paris

This course is organized by François Pacaud and Armin Nurkanovic. Lectures will be given by Armin Nurkanovic. This three-day course aims to provide both theoretical background and hands-on practical knowledge in formulating optimal control problems subject to hybrid systems with switches and state jumps. Participants will learn to:

Formulate and solve optimal control problems for hybrid systems
Master specialized numerical methods for hybrid system optimization
Apply these concepts to real-world engineering problems

Lecture slides

Theory and algorithms for nonlinear programming - (slides, recording)
Direct methods for smooth nonlinear optimal control - (slides, recording)
Introduction to nonsmooth differential equations and hybrid systems - (slides, recording)
Direct methods for nonsmooth nonlinear optimal control - (slides, recording)
MPECs: nonsmooth modelling and optimality conditions - (slides, recording)
Numerical methods for mathematical programs with complementarity constraints - (slides, recording)

→ Read more...

2025/06/04 22:07 · Horea Caramizaru · 0 Comments · 0 Linkbacks

francois-pacaud, and, armin-nurkanovic, nonsmooth-dynamics, mpc, control, optimization, contact-dynamics, mines-paris-psl, uni-freiburg, 2025

Linear Programming

Course Playlist
Class homepage
Course notes
Recomanded book

→ Read more...

2024/12/10 18:33 · Horea Caramizaru · 0 Comments · 0 Linkbacks

optimization, polytopes, integer-linear-programming, convex-polyhedra, simplex-method, duality, farkas-lemma, ellipsoid-method, max-flow-min-cut, optimal, transport, kantarovich-rubenstein, branch-and-bound, dynamic-programming, linear-programming, colorado-state-university, henry-adams, 2020

ChE-597-Computational-Optimization

Playlist course

→ Read more...

2024/12/10 14:05 · Horea Caramizaru · 0 Comments · 0 Linkbacks

optimization, nlp, linear-programming, purdue-university, can-li, 2024

Algorithms and Architectures for Improving Optimization Layers

→ Read more...

2024/12/03 23:02 · Horea Caramizaru · 0 Comments · 0 Linkbacks

slam, optimization, robotics, cmu, swaminathan-gurumurthy, 2024

How to Calculate Derivative of Vectors and Matrices!

→ Read more...

2024/12/02 10:17 · Horea Caramizaru · 0 Comments · 0 Linkbacks

derivateves, optimization, 2024

Leveraging Parallelism to Accelerate Quadratic Program Solvers for MPC

→ Read more...

2024/11/21 14:38 · Horea Caramizaru · 0 Comments · 0 Linkbacks

cmu, robotics, optimization, mpc, whole-body-mpc, qp, quadratic-programming, alternating-direction-method-of-multipliers, admm, relu, osqp, proxqp, arun-bishop, 2024

Leveraging Structure for Efficient and Dexterous Contact-Rich Manipulation

→ Read more...

2024/11/12 17:29 · Horea Caramizaru · 0 Comments · 0 Linkbacks

hyung-ju-terry-suh, contact-dynamics, robotics, optimization, phd-thesis, mit, 2024

Advancing Legged Robot Agility: from Video Imitation to GPU Acceleration

Abstract:

Achieving human and animal-level agility has been a long-standing goal in robotics research. Recent advancements in numerical optimization and machine learning have pushed legged systems to greater capabilities than ever before, enabling black flips, parkour, and manipulation of heavy objects. Despite these exciting developments, this thesis identifies two key limitations of current legged robot technology and aims to improve upon existing art.

First, legged robots today require manual specifications of desired behaviors and fail to learn from their human and animal counterparts. We introduce SLoMo, a first-of-its-kind framework for transferring skilled motions from casually captured videos of humans and animals to legged robots. From a monocular RGB video, SLoMo synthesizes physically plausible trajectories for downstream offline trajectory optimization and online predictive control of quadruped or humanoid robots. We demonstrate SLoMo by transferring cat and dog motions to quadruped robot hardware and human motions to a simulated humanoid robot.

Second, current model-predictive control (MPC) for legged systems often resort to simplified models due to computational limitations in real-time settings. This is due to the high dimensionality of these robots and the reliance of existing numerical optimization algorithms on fundamentally serial, CPU-friendly linear algebra routines. We leverage advancements in GPU parallelization by developing a quadratic programming (QP) solver that uses only GPU-friendly operations. We refer to our solver as ReLU-QP, thanks to its computational similarities to inferencing a deep neural network with rectified linear unit (ReLU) activation functions. Across benchmarks on solving random QPs and high-dimensional MPC tasks in simulation, including balancing a full-order Atlas humanoid robot on one foot under control limits, ReLU-QP shows an order-of-magnitude speed improvement over state-of-the-art CPU-based QP solvers and solves MPC for modern legged robots at kilohertz rates.

→ Read more...

2024/11/08 10:13 · Horea Caramizaru · 0 Comments · 0 Linkbacks

optimization, control, robotics, cmu, john-zhang, 2024

Policy gradient for black-box optimization

Policy gradient method are widely used in the Reinforcement Learning settings. In this post we build policy gradient from the ground up, starting from the easier static scenario first, where we maximize a reward function {r} depending solely on our control variable {x}. In subsequent posts, we will turn our attention to the contextual bandit setting, where the reward also depends on a “state” that evolves. Finally, we will turn to the “full-blown” Reinforcement Learning scenario, where state evolves endogenously, as a function of the control variable.

Blogpost Link

→ Read more...

2024/11/08 10:04 · Horea Caramizaru · 0 Comments · 0 Linkbacks

policy-gradient, optimization, lorenzo-maggi, 2023

Lagrangian multipliers, normal cones and KKT optimality conditions

Blogpost Link

→ Read more...

2024/11/08 09:54 · Horea Caramizaru · 0 Comments · 0 Linkbacks

lagrangian-multipliers, normal-cones, kkt, optimality-conditions, optimization, lorenzo-maggi, 2024

Game physics

Full Playlist

→ Read more...

2024/11/05 21:38 · Horea Caramizaru · 0 Comments · 0 Linkbacks

game-development, physics-simulation, optimization, 2024

Co-Design Optimization of Morphing Topology and Control of Winged Drones

Abstract

The design and control of winged aircraft and drones is an iterative process aimed at identifying a compromise of mission-specific costs and constraints. When agility is required, shape-shifting (morphing) drones represent an efficient solution. However, morphing drones require the addition of actuated joints that increase the topology and control coupling, making the design process more complex. We propose a co-design optimization method that assists the engineers by proposing a morphing drone’s conceptual design that includes topology, actuation, morphing strategy, and controller parameters. The method consists of applying multi-objective constraint-based optimization to a multi-body winged drone with trajectory optimization to solve the motion intelligence problem under diverse flight mission requirements, such as energy consumption and mission completion time. We show that co-designed morphing drones outperform fixed-winged drones in terms of energy efficiency and mission time, suggesting that the proposed co-design method could be a useful addition to the aircraft engineering toolbox.

→ Read more...

2024/05/27 08:12 · Horea Caramizaru · 0 Comments · 0 Linkbacks

co-design, optimization, morphing-topology, mpc, fabio-bergonti, gabriele-nava, valentin-wueest, antonello-paolino, giuseppe-l erario, daniele-pucci, dario-floreano, iit, university-of-manchester, epfl, university-of-naples-federico-ii, 2024

Alice's Adventures in a Differentiable Wonderland -- Volume I, A Tour of the Land

Abstract

This book is a self-contained introduction to the design of modern (deep) neural networks. Because the term “neural” comes with a lot of historical baggage, I prefer the simpler term “differentiable models” in the text. The focus of this 250-pages volume is on building efficient blocks for processing nD data, including convolutions, transformers, graph layers, and modern recurrent models (including linearized transformers and structured state-space models). Because the field is evolving quickly, I have tried to strike a good balance between theory and code, historical considerations and recent trends. I assume the reader has some exposure to machine learning and linear algebra, but I try to cover the preliminaries when necessary. The volume is a refined draft from a set of lecture notes for a course called Neural Networks for Data Science Applications that I teach in Sapienza. I do not cover many advanced topics (generative modeling, explainability, prompting, agents), which will be published over time in the companion website.

Arxiv Link
Book Link

→ Read more...

2024/05/02 13:59 · Horea Caramizaru · 0 Comments · 0 Linkbacks

math, ml, ad, optimization, simone-scardapane, 2024

FEM@LLNL | Supercharging Programming Through Compiler Technology

→ Read more...

2024/04/29 23:35 · Horea Caramizaru · 0 Comments · 0 Linkbacks

math, optimization, ad, compilers-design, llvm, enzyme, william-moses, illinois-urbana-champaign, 2024

Inexact Newton with Iterated Sensitivities for Optimization of Partial Differential Equation

PhD Theses link

Presentation

→ Read more...

2024/04/26 08:45 · Horea Caramizaru · 0 Comments · 0 Linkbacks

math, numerics, optimization, control, pde, newton, sensitivities, justin-asbjorn-pearse-danker, uni-freiburg, 2020

TCOptRob Seminar: Learning complex behaviors with nonlinear MPC by Ludovic Righetti of NYU

→ Read more...

2024/04/03 21:09 · Horea Caramizaru · 0 Comments · 0 Linkbacks

tcoptrob, model-based-optimization, mpc, ludovic-righetti, nyu, robotics, optimization, rl, sqp, ddp, lqr, kkt, tri-diagonal-kkt, thomas-algorithm, crocoddyl, 2024

Phebe Vayanos - Integer optimization for predictive & prescriptive analytics in high stakes domains

→ Read more...

2024/04/03 08:18 · Horea Caramizaru · 0 Comments · 0 Linkbacks

phebe-vayanos, university-of-southern-california, integer-optimization, optimization, robust-optimization, ipam, 2023

Optimization for Control and Planning of Multi-contact Dynamic Motion

→ Read more...

2024/04/01 02:35 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mit, robotics, optimization, control, planning, multi-contact, michael-posa, contact-implicit-mpc, 2017

Distributed Optimization via Alternating Direction Method of Multipliers

→ Read more...

2024/03/31 20:06 · Horea Caramizaru · 0 Comments · 0 Linkbacks

distributed-optimization, admm, optimization, stephen-boyd, microsoft-research, 2016

Matrix Calculus for Machine Learning and Beyond

Course Description

→ Read more...

2024/03/24 15:19 · Horea Caramizaru · 0 Comments · 0 Linkbacks

math, optimization, derivatives, automatic-differentiation, algorithmic-differentiation, ml, alan-edelman, steven-johnson, mit, mit-opencourseware, 2023, course

The Elements of Differentiable Programming

Abstract

Arxiv Link

→ Read more...

2024/03/24 14:35 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mathieu-blondel, vincent-roulet, optimization, derivatives, automatic-differentiation, algorithmic-differentiation, differentiable-programming, ml

Robust Optimization

→ Read more...

2024/02/19 22:46 · Horea Caramizaru · 0 Comments · 0 Linkbacks

phebe-vayanos, robust-optimization, sequential-decision-making, optimization

Optimal Control 2023

→ Read more...

2023/12/24 16:20 · Horea Caramizaru · 0 Comments · 0 Linkbacks

Optimal Control - 2022

→ Read more...

2023/12/24 16:00 · Horea Caramizaru · 0 Comments · 0 Linkbacks

Linear Programming

→ Read more...

2023/12/23 21:41 · Horea Caramizaru · 0 Comments · 0 Linkbacks

course, math, math484, lp, simplex, optimization, penn-state-university, wen-shen, 2016

Optimization

→ Read more...

2023/12/23 21:32 · Horea Caramizaru · 0 Comments · 0 Linkbacks

course, math, numerics, sqp, qp, lp, optimization, simplex

Contact and Friction Simulation for Computer Graphics - 2022

→ Read more...

2023/12/22 15:23 · Horea Caramizaru · 0 Comments · 0 Linkbacks

siggraph, course, contact, friction, simulation, computer-graphics, optimization, lcp, sheldon-andrews, kenny-erleben, zachary-ferguson, 2022

Contact and Friction Simulation for Computer Graphics - 2021

→ Read more...

2023/12/22 15:17 · Horea Caramizaru · 0 Comments · 0 Linkbacks

siggraph, course, contact, friction, simulation, computer-graphics, optimization, lcp, sheldon-andrews, kenny-erleben, 2021

Optimization on Manifolds (course)

Playlist Link

Abstract:

Smooth optimization on manifolds naturally generalizes smooth optimization in Euclidean spaces in a manner that is of interest in a variety of applications, including modal analysis, blind source separation (via independent component analysis), pose estimation in computer vision, and model reduction in dynamical systems. Manifolds of interest include the Stiefel manifolds and Grassmann manifolds.

After presenting a number of motivating applications, we will introduce the basics of differential manifolds and Riemannian geometry, and describe how methods in optimization, like line-search, Newton's method, and trust region methods can be generalized to the case of manifolds. The course will assume only a basic knowledge of matrix algebra and real analysis.

→ Read more...

2023/12/04 16:59 · Horea Caramizaru · 0 Comments · 0 Linkbacks

course, math, melvin-leok, optimization, manifolds

Controlled Sparsity via Constrained Optimization

→ Read more...

2023/11/24 22:13 · Horea Caramizaru · 0 Comments · 0 Linkbacks

sparsity, optimization, regularization, jose-gallego-posada

Differentiable Simulation Methods for Robotic Agent Design

→ Read more...

2023/11/24 20:34 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mit, robotics, differentiable-simulation, optimization

So You Think You Know How to Take Derivatives? | Steven Johnson

→ Read more...

2023/11/24 20:14 · Horea Caramizaru · 0 Comments · 0 Linkbacks

optimization, derivatives, automatic-differentiation, algorithmic-differentiation

Differentiable Simulation Methods for Robotic Agent Design

https://people.iiis.tsinghua.edu.cn/~taodu/

Thesis: "Differentiable Simulation Methods for Robotic Agent Design" by Tao Du

Abstract:

Designing robots with extreme performance in a given task has long been an exciting research problem drawing attention from researchers in robotics, graphics, and artificial intelligence. As a robot is a combination of its hardware and software, an optimal robot requires both an excellent implementation of its hardware (e.g., morphological, topological, and geometrical designs) and an outstanding design of its software (e.g., perception, planning, and control algorithms). While we have seen promising breakthroughs for automating a robot's software design with the surge of deep learning in the past decade, exploration of optimal hardware design is much less automated and is still mainly driven by human experts, a process that is both labor-intensive and error-prone. Furthermore, experts typically optimize a robot's hardware and software separately, which may miss optimal designs that can only be revealed by optimizing its hardware and software simultaneously. This thesis argues that it is time to rethink robot design as a holistic process where a robot's body and brain should be co-optimized jointly and automatically. In this thesis, we present a computational robot design pipeline with differentiable simulation as a key player. We first introduce the concept of computational robot design on a real-world copter whose geometry and controller are co-optimized with a differentiable simulator, resulting in a custom copter that outperforms designs suggested by human experts by a substantial margin. Next, we push the boundary of differentiable simulation by developing advanced differentiable simulators for soft-body and fluid dynamics. Contrary to traditional belief, we show that deriving gradients for such intricate, high-dimensional physics systems can be both science and art. Finally, we discuss challenges in transferring computational designs discovered in simulation to real-world hardware platforms. We present a solution to this simulation-to-reality transfer problem using our differentiable simulator on an example of modeling and controlling a real-world soft underwater robot. We conclude this thesis by discussing open research directions in differentiable simulation and envisioning a fully automated computational design pipeline for real-world robots in the future.

→ Read more...

2023/11/02 11:36 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mit, phd-thesis, differentiable-simulation, optimization, robotics, soft-body, fluid, dynamics, computational-design, simulation-to-reality, underwater-robot, computational-design-pipeline, tao-du, 2021

Optimization on Manifolds

→ Read more...

2023/11/01 23:08 · Horea Caramizaru · 0 Comments · 0 Linkbacks

optimization, manifolds

Thesis Scientific Computing

2022, the year when I defended my Thesis, “Multi-body modeling of robot dynamics and system identification during MPC”, in Scientific Computing. The updated translation of the poem can be found here.

Abstract

Due to external influences over parameters that characterize dynamical systems, an online parameter estimation must be added as part of model predictive control strategies. In this thesis, we show how continuous parameters estimation, using inverse dynamics, can be used for identifying the inertial parameters (mass, inertia, and center of mass) of multi-body systems as part of an adaptive control strategy. For this, a Featherstone spatial algebra equivalent model, based on screw theory was used. The system identification was done using a linear least squares approach using the Recursive Newton-Euler Algorithm as a way of implementing a generic solution. The process is for open-loop robots and is tested using an optimal control algorithm based on multiple shooting.

→ Read more...

2023/08/25 23:49 · Horea Caramizaru · 0 Comments · 0 Linkbacks

msc, thesis, mpc, optimal-control, optimization, nlp, parametric-estimation, ode, dae, multibody-dynamics, featherstone, spatial-algebra, casadi, robotics, 2022

Robotics

anonymous@undisclosed.example.com (Anonymous) — Fri, 24 Nov 2023 16:34:49 +0000

Robotics

Robot Learning 2025: Foundational Models for Robotics and Scaling DeepRL

Course playlist

→ Read more...

2026/04/18 17:41 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, glen-berseth, rl, deeprl, foundational-models, vlm, vla, behavior-cloning, universite-de-montreal, 2025

Introduction to Flow Matching and Diffusion Models

MIT Computer Science Class 6.S184: Generative AI with Stochastic Differential Equations

Diffusion and flow-based models have become the state of the art for generative AI across a wide range of data modalities, including images, videos, shapes, molecules, music, and more! This course aims to build up the mathematical framework underlying these models from first principles. At the end of the class, students will have built a toy image diffusion model from scratch, and along the way, will have gained hands-on experience with the mathematical toolbox of stochastic differential equations that is useful in many other fields. This course is ideal for students who want to develop a principled understanding of the theory and practice of generative AI.

Course Notes

The course notes serve as the backbone of the course and provide a self-contained explanation of all material in the class. In contrast, lecture slides will generally not be self-contained and are intended to provide accompanying visualizations during the lecture. You may view the notes by clicking on the colored link below.

Course Notes

The full video course and slides/course materials can be found here:

Course web page

2025/11/17 23:06 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, ml, rl, diffusion-models, stochastic-differential-equations, sde, mit, 2025

Optimization and Learning for Robot Control 2025

→ Read more...

2025/11/17 22:38 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, lqr, ilqr, task-space-control, ddp, ml, optimal-control, mpc, optimization, qp, andrea-del-prete, 2025

Memmo Winter School 2019

Here is a list of topics that will be covered during this school

Movement primitives
Pinocchio: kinematics and derivatives in C++ and python
Optimal control solvers
Contact planner
Walking robot navigation and state estimation
Motion planning for legged locomotion (tentative title, probably focus on quadrupedal locomotion)
Motion planning workflow for the atalante exoskeleton
TALOS control system architecture and whole body controller
Deep RL
QP-based task-space inverse
Dynamics: theory and implementation
Robust optimization with application to task-space inverse dynamics

→ Read more...

2025/11/17 17:41 · Horea Caramizaru · 0 Comments · 0 Linkbacks

Summer School on Direct Methods for Optimal Control of Nonsmooth Systems

This four-day intensive course aims to provide both theoretical background and hands-on practical knowledge in formulating and numerical methods to solve optimal control problems with nonsmooth differential equation models with switches and state jumps. Nonsmooth dynamical systems arise in robotics, chemical engineering, biology, mechatronics, or aerospace, as soon as some if-else statements, switches, and state jump are encoded in the systems’ dynamics. For example, contacts and friction in robotic systems lead to jumps and switches.

→ Read more...

2025/09/30 14:01 · 0 Comments · 0 Linkbacks

summer-school, optimal-control, nonsmooth-systems, direct-methods, filippov-systems, dae, direct-collocation, hybrid-systems, elastic-impacts, fesd, mpcc, freiburg-university, robotics, 2023

Real-time constrained trajectory optimization in robotics: theory, implementation, and application

Abstract

→ Read more...

2025/06/09 10:53 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, optimal-control, optimization, mpc, wilson-jallet, inria, cnrs, phd, 2024

DeepMind x UCL | Introduction to Reinforcement Learning 2015

Full course

→ Read more...

2025/04/24 15:05 · Horea Caramizaru · 0 Comments · 0 Linkbacks

learning, rl, robotics, control, david-silver, ucl, 2015

Algorithms and Architectures for Improving Optimization Layers

→ Read more...

2024/12/03 23:02 · Horea Caramizaru · 0 Comments · 0 Linkbacks

slam, optimization, robotics, cmu, swaminathan-gurumurthy, 2024

Stanford Seminar - The Trouble with Contact: Helping Robots Touch the World

→ Read more...

2024/12/03 23:01 · Horea Caramizaru · 0 Comments · 0 Linkbacks

stanford, robotics, contact-dynamics, cmu, aaron-johnson, 2024

Leveraging Parallelism to Accelerate Quadratic Program Solvers for MPC

→ Read more...

2024/11/21 14:38 · Horea Caramizaru · 0 Comments · 0 Linkbacks

cmu, robotics, optimization, mpc, whole-body-mpc, qp, quadratic-programming, alternating-direction-method-of-multipliers, admm, relu, osqp, proxqp, arun-bishop, 2024

Leveraging Structure for Efficient and Dexterous Contact-Rich Manipulation

→ Read more...

2024/11/12 17:29 · Horea Caramizaru · 0 Comments · 0 Linkbacks

hyung-ju-terry-suh, contact-dynamics, robotics, optimization, phd-thesis, mit, 2024

Advancing Legged Robot Agility: from Video Imitation to GPU Acceleration

Abstract:

→ Read more...

2024/11/08 10:13 · Horea Caramizaru · 0 Comments · 0 Linkbacks

optimization, control, robotics, cmu, john-zhang, 2024

Control Architectures for Agile-Legged Robots

→ Read more...

2024/10/21 06:46 · Horea Caramizaru · 0 Comments · 0 Linkbacks

motion-planning, whole-body-control, joint-control, actuators, robotics, georgia-tech, donghyun-kim, daros-lab, 2021

Autonomy Talks - Stelian Coros: Physics-based Modeling and the Quest for Intelligent Robots

→ Read more...

2024/06/04 13:10 · Horea Caramizaru · 0 Comments · 0 Linkbacks

physics-based-modeling, reality-gap, differentiable-simulation, learning, robotics, stelian, coros, eth, 2022

Rethinking Robustness Assessment: Adversarial Attacks on Learning-based Quadrupedal Locomotion Controllers

→ Read more...

2024/06/03 08:55 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, mpc, adversarial-attacks, learning, quadruped, locomotion, fan-shi, stelian-coros, eth, 2024

PhD Thesis Defense - Sangwoon Kim - TEXterity: Tactile Extrinsic deXterity

Abstract:

This thesis introduces the concept of TEXterity (Tactile Extrinsic deXterity) to address challenges in robotic manipulation. Focusing on tactile sensing, TEXterity aims to enhance dexterity by enabling robots to perceive and act upon extrinsic contact between the grasped object and the environment. Identifying interpretability, observability, and uncertainty as key challenges in tactile sensing, this thesis sets out to answer four pivotal questions:

Is tactile sensing actually useful?
How can we exploit tactile sensing efficiently?
How can we reason about extrinsic contact with tactile sensing?
How can we achieve extrinsic dexterity with tactile sensing?

The conclusion summarizes the key findings, emphasizing the significance of tactile sensing and TEXterity in addressing challenges and advancing robotic manipulation. Strategies to tackle major challenges are outlined, focusing on interpretability, observability, and uncertainty. In essence, this thesis lays the groundwork for unlocking the potential of tactile sensing in robotic manipulation, offering insights, solutions, and avenues for future research to propel the field toward achieving TEXterity and further toward human-level dexterity.

→ Read more...

2024/05/03 20:39 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, phd-thesis, tactile, sensing, dexterity, sangwoon-kim, mit, 2024

Jemin Hwangbo (KAIST): Large-scale policy training for robots

→ Read more...

2024/04/25 22:02 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, jemin-hwangbo, kaist, rl

TCOptRob Seminar: Learning complex behaviors with nonlinear MPC by Ludovic Righetti of NYU

→ Read more...

2024/04/03 21:09 · Horea Caramizaru · 0 Comments · 0 Linkbacks

tcoptrob, model-based-optimization, mpc, ludovic-righetti, nyu, robotics, optimization, rl, sqp, ddp, lqr, kkt, tri-diagonal-kkt, thomas-algorithm, crocoddyl, 2024

Optimization for Control and Planning of Multi-contact Dynamic Motion

→ Read more...

2024/04/01 02:35 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mit, robotics, optimization, control, planning, multi-contact, michael-posa, contact-implicit-mpc, 2017

Towards Computational Design of Shape and Control for Rigid Robots

→ Read more...

2024/04/01 02:06 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mit, computational-design, robotics, rigid-robots, jie-xu, 2022

MIT Robotics - Dieter Fox - Toward Foundational Robot Manipulation Skills

→ Read more...

2024/03/31 18:04 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mit, mit-robotics, robotics, dieter-fox, robot, manipulation, 2023

Princeton Robotics - Russ Tedrake - Dexterous Manipulation with Diffusion Policies

→ Read more...

2024/03/31 14:38 · Horea Caramizaru · 0 Comments · 0 Linkbacks

princeton, princeton-robotics, robotics, irom-lab, russ-tedrake, dexterous-manipulation, diffusion-policies, 2023

Building an Internal Cycloidal Robotic Actuator

→ Read more...

2024/03/09 08:59 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, actuators, diy, mechatronics

Advancements in Trajectory Optimization and Model Predictive Control for Legged Systems

- 2nd Edition -
2024 IEEE International Conference on Robotics and Automation in PACIFICO Yokohama

Robotic technology has proven to be an excellent solution for aiding humans in an ever-increasing number of scenarios: from domestic and urban routines to industrial tasks, robots reduce the workload burden on humans and their exposure to hazards. Despite the capabilities demonstrated in recent years, many obstacles remain to be overcome. New challenges arise from the increasing capabilities of advanced robotic platforms. The more complex and unstructured the environment, the more robots should be versatile and reliable to overcome obstacles, plan optimal motions, and reliably accomplish the designed tasks.

→ Read more...

2024/02/11 21:59 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, trajectory-optimization, mpc, legged, yokohama, 2024

Optimal Control 2023

→ Read more...

2023/12/24 16:20 · Horea Caramizaru · 0 Comments · 0 Linkbacks

Optimal Control - 2022

→ Read more...

2023/12/24 16:00 · Horea Caramizaru · 0 Comments · 0 Linkbacks

Computational Symmetry and Learning for Robotics

→ Read more...

2023/12/24 09:27 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mit, mit-robotics, robotics, computational, symmetry, learning, lie-group, lie-algebra, maani-ghaffari, university-of-michigan, 2023

Robust Autonomous Vehicle Localization using GPS: from Tandem Drifting Cars to “GPS” on the Moon

→ Read more...

2023/12/24 09:10 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mit, robotics, mit-robotics, grace-gao, localization, gps, stanford-university, 2023

Computational Design of Origami and Compliant Robots

→ Read more...

2023/12/24 09:01 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mit, robotics, mit-robotics, cynthia-sung, computational, design, origami, compliant, university-of-pennsylvania, 2023

Differentiable Simulation Methods for Robotic Agent Design

→ Read more...

2023/11/24 20:34 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mit, robotics, differentiable-simulation, optimization

Geometric mechanics of continuum and articulated soft robots

→ Read more...

2023/11/24 20:25 · Horea Caramizaru · 0 Comments · 0 Linkbacks

mechanics-of-continuum, lie-group, lie-algebra, soft-robots, robotics, stanislao-grazioso

2.160 Identification, Estimation, and Learning

Playlist Link

Prof. Harry Asada Massachusetts Institute of Technology Department of Mechanical Engineering Fall 2020 All of the lecture recordings, slides, and notes are available on our lab website: External Link

→ Read more...

2023/11/18 23:25 · Horea Caramizaru · 0 Comments · 0 Linkbacks

course, math, mit, identification, estimation, learning, dlabroboticsmit, ekf, ukf, particle-filter, robotics

Combining learning and geometry for robot perception and navigation

→ Read more...

2023/11/08 09:56 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, perception, navigation, ml, geometry, maani-ghaffari, may-2022

Motion Planning Around Obstacles with Graphs of Convex Sets

→ Read more...

2023/11/06 17:18 · Horea Caramizaru · 0 Comments · 0 Linkbacks

ri-seminar, russ-tedrake, robotics, motion-planning, cmu-robotics, 2023

From Compliant Mechanisms to Hyper-Elastic Robots

→ Read more...

2023/11/05 10:49 · Horea Caramizaru · 0 Comments · 0 Linkbacks

ri-seminar, michael-wang, robotics, cmu-robotics-institute, compliant-mechanisms, compliant, hyper-elastic-robots, 2014

Do Differentiable Simulators Give Better Policy Gradients?

"Do Differentiable Simulators Give Better Policy Gradients?" by H.J. Terry Suh, Max Simchowitz, Kaiqing Zhang, Russ Tedrake

Abstract:

Differentiable simulators promise faster computa- tion time for reinforcement learning by replacing zeroth-order gradient estimates of a stochastic objective with an estimate based on first-order gradients. However, it is yet unclear what fac- tors decide the performance of the two estimators on complex landscapes that involve long-horizon planning and control on physical systems, despite the crucial relevance of this question for the util- ity of differentiable simulators. We show that characteristics of certain physical systems, such as stiffness or discontinuities, may compromise the efficacy of the first-order estimator, and ana- lyze this phenomenon through the lens of bias and variance. We additionally propose an α-order gra- dient estimator, with α ∈ [0, 1], which correctly utilizes exact gradients to combine the efficiency of first-order estimates with the robustness of zero- order methods. We demonstrate the pitfalls of traditional estimators and the advantages of the α-order estimator on some numerical examples.

→ Read more...

2023/11/02 11:55 · Horea Caramizaru · 0 Comments · 0 Linkbacks

robotics, differentiable-simulators, reinforcement-learning