Carlos Mastalli

Carlos Mastalli
Heriot-Watt University

PhD in Bioengineering and Robotics

About

41
Publications
17,194
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
1,069
Citations
Citations since 2017
32 Research Items
1049 Citations
2017201820192020202120222023050100150200250300
2017201820192020202120222023050100150200250300
2017201820192020202120222023050100150200250300
2017201820192020202120222023050100150200250300
Introduction
I am a robotics researcher in the working on the topic of multi-contact planning and control, and machine learning in legged robots. My main motivation is to enable robots to move everywhere. In particular, I am focused on legged locomotion since it combines all the main challenges in robot mobility: high-dimensional and unactuated systems, discontinuous dynamics, robot interaction and perception. For that, I am interested in methods that combine optimal control and machine learning, allowing us to explore the knowledge of the system but also the vast robot’s data. I completed my PhD on “Planning and Execution of Dynamic Whole-Body Locomotion on Challenging Terrain” in April 2017 in the Dynamic Legged System lab at Istituto Italiano di Tecnologia. After that, I served as Research Fell

Publications

Publications (41)
Preprint
Full-text available
We introduce Crocoddyl (Contact RObot COntrol by Differential DYnamic Library), an open-source framework tailored for efficient multi-contact optimal control. Crocoddyl efficiently computes the state trajectory and the control policy for a given predefined sequence of contacts. Its efficiency is due to the use of sparse analytical derivatives, expl...
Preprint
Full-text available
Legged robots promise an advantage over traditional wheeled systems, however, most legged robots are still confined to structured and flat environments. One of the main reasons for this is the difficulty in planning complex whole-body motions while taking into account the terrain conditions. This problem is very high-dimensional as it considers the...
Conference Paper
Full-text available
We present a trajectory optimization framework for locomotion on rough terrain. We jointly optimize the Center of Mass (CoM) motion and the foothold locations, while considering terrain conditions. We use a terrain costmap to quantify the desirability of a foothold location. We increase the gait’s adaptability to the terrain by optimizing the step...
Conference Paper
Full-text available
Most animal and human locomotion behaviors for solving complex tasks involve dynamic motions and rich contact interaction. In fact, complex maneuvers need to consider dynamic movement and contact events at the same time. We present a hierarchical trajectory optimization approach for planning dynamic movements with unscheduled contact sequences. We...
Article
Full-text available
Traditional motion planning approaches for multi-legged locomotion divide the problem into several stages, such as contact search and trajectory generation. However, reasoning about contacts and motions simultaneously is crucial for the generation of complex whole-body behaviors. Currently, coupling theses problems has required either the assumptio...
Article
Full-text available
Differential dynamic programming (DDP) is a direct single shooting method for trajectory optimization. Its efficiency derives from the exploitation of temporal structure (inherent to optimal control problems) and explicit roll-out/integration of the system dynamics. However, it suffers from numerical instability and, when compared to direct multipl...
Preprint
Full-text available
Optimal control (OC) using inverse dynamics provides numerical benefits such as coarse optimization, cheaper computation of derivatives, and a high convergence rate. However, in order to take advantage of these benefits in model predictive control (MPC) for legged robots, it is crucial to handle its large number of equality constraints efficiently....
Preprint
Full-text available
Robot design optimization, imitation learning and system identification share a common problem which requires optimization over robot or task parameters at the same time as optimizing the robot motion. To solve these problems, we can use differentiable optimal control for which the gradients of the robot's motion with respect to the parameters are...
Preprint
Full-text available
Achieving agile maneuvers through multiple contact phases has been a longstanding challenge in legged robotics. It requires to derive motion plans and local control feedback policies in real-time to handle the nonholonomy of the kinetic momenta. While a few recent predictive control approaches based on centroidal momentum have been able to generate...
Preprint
Full-text available
We present a versatile framework for the computational co-design of legged robots and dynamic maneuvers. Current state-of-the-art approaches are typically based on random sampling or concurrent optimization. We propose a novel bilevel optimization approach that exploits the derivatives of the motion planning sub-problem (i.e., the lower level). The...
Preprint
Full-text available
Last decades of humanoid research has shown that humanoids developed for high dynamic performance require a stiff structure and optimal distribution of mass--inertial properties. Humanoid robots built with a purely tree type architecture tend to be bulky and usually suffer from velocity and force/torque limitations. This paper presents a novel seri...
Preprint
Full-text available
Differential dynamic programming (DDP) is a direct single shooting method for trajectory optimization. Its efficiency derives from the exploitation of temporal structure (inherent to optimal control problems) and explicit roll-out/integration of the system dynamics. However, it suffers from numerical instability and, when compared to direct multipl...
Conference Paper
Full-text available
An important issue when synthesizing legged loco-motion plans is the combinatorial complexity that arises from gait pattern selection. Though it can be defined manually, the gait pattern plays an important role in the feasibility and optimality of a motion with respect to a task. Replacing human intuition with an automatic and efficient approach fo...
Conference Paper
Full-text available
Recently, Differential Dynamic Programming (DDP) and other similar algorithms have become the solvers of choice when performing non-linear Model Predictive Control (nMPC) with modern robotic devices. The reason is that they have a lower computational cost per iteration when compared with off-the-shelf Non-Linear Programming (NLP) solvers, which ena...
Article
Full-text available
Planning whole-body motions while taking into account the terrain conditions is a challenging problem for legged robots since the terrain model might produce many local minima. Our coupled planning method uses stochastic and derivatives-free search to plan both foothold locations and horizontal motions due to the local minima produced by the terrai...
Conference Paper
Full-text available
We introduce Crocoddyl (Contact RObot COntrol by Differential DYnamic Library), an open-source framework tailored for efficient multi-contact optimal control. Crocoddyl efficiently computes the state trajectory and the control policy for a given predefined sequence of contacts. Its efficiency is due to the use of sparse analytical derivatives, expl...
Article
Full-text available
We present experimental results using a passive whole-body control approach for quadruped robots that achieves dynamic locomotion while compliantly balancing the robot’s trunk. We formulate the motion tracking as a Quadratic Program(QP) that takes into account the full robot rigid body dynamics,the actuation limits, the joint limits and the contact...
Conference Paper
Full-text available
In this paper, we propose a framework to build a memory of motion for warm-starting an optimal control solver for the locomotion task of a humanoid robot. We use HPP Loco3D, a versatile locomotion planner, to generate offline a set of dynamically consistent whole-body trajectory to be stored as the memory of motion. The learning problem is formulat...
Preprint
Full-text available
In this paper, we propose a framework to build a memory of motion to warm-start an optimal control solver for the locomotion task of the humanoid robot Talos. We use HPP Loco3D, a versatile locomotion planner, to generate offline a set of dynamically consistent whole-body trajectory to be stored as the memory of motion. The learning problem is form...
Conference Paper
Full-text available
This paper describes the implementation of a motion generation pipeline motivated by vision for a TALOS humanoid robot. From a starting configuration and given a set of visual features and their desired values, the problem is to find a motion which makes the robot reach the desired values of the visual features. In order to achieve a feasible Gazeb...
Preprint
Full-text available
We present experimental results using a passive whole-body control approach for quadruped robots that achieves dynamic locomotion while compliantly balancing the robot’s trunk. We formulate the motion tracking as a Quadratic Program (QP) that takes into account the full robot rigid body dynamics, the actuation limits, the joint limits and the conta...
Conference Paper
Full-text available
A common strategy to generate efficient locomo-tion movements is to split the problem into two consecutive steps: the first one generates the contact sequence together with the centroidal trajectory, while the second step computes the whole-body trajectory that follows the centroidal pattern. While the second step is generally handled by a simple p...
Chapter
Full-text available
The quality of visual feedback can vary signif- icantly on a legged robot meant to traverse unknown and unstructured terrains. The map of the environment, acquired with online state-of-the-art algorithms, often degrades after a few steps, due to sensing inaccuracies, slippage and unexpected disturbances. If a locomotion algorithm is not designed to...
Preprint
Full-text available
The quality of the visual feedback can vary significantly on a legged robot that is meant to traverse unknown and unstructured terrains. The map of the environment, acquired with online state-of-the-art algorithms, often degrades after a few steps, due to sensing inaccuracies, slippage and unexpected disturbances. When designing locomotion algorith...
Data
file also available at: https://www.youtube.com/watch?v=vUx5b5kfRfE&feature=youtu.be
Article
Full-text available
Motion planning in multi-contact scenarios has recently gathered interest within the legged robotics community, however actuator force/torque limits are rarely considered. We believe that these limits gain paramount importance when the complexity of the terrains to be traversed increases. We build on previous research from the field of robotic gras...
Poster
Full-text available
Planning motions for multi-legged robots is a complex task, as it involves both the discrete choice on gait sequence, and continuous decisions on foot location and robot dynamics. For this reason, most approaches have decoupled the problem in two stages: searching contacts and planning motions. However, by ignoring the dynamics in the first stage,...
Poster
Full-text available
This is the content of the poster I will present at the IROS 2017 workshop ”Frontiers in Contact-rich Robotic Interaction: Modeling, Optimization and Control Synthesis” in Vancouver. The idea is to devise a feasibility metrics that keeps both stability and actuation consistency into account. Coupled with the centroidal dynamic model of the robot, t...
Thesis
Full-text available
Legged vehicles present a potential advantage over traditional wheeled systems since they offer greater mobility in rough and challenging terrain. However, most legged robots are still confined to structured and flat terrain. One of the main reasons for this is the difficulty in planning complex whole-body motions while taking into account future t...
Conference Paper
Full-text available
This work presents a new formulation of a modular relative Jacobian used to control combined manipulators as a single manipulator with a single effector. In particular, this modular relative Jacobian is designed for 3-arm parallel manipulators. It is called a relative Jacobian because it is expressed relative to the reference frames at the manipula...
Chapter
Full-text available
In this work is developed an architecture which consists of four main components: perception system, tasks planning, motion planner, and control systems that allow autonomous operations in backhoe machines. In the first part is described the architecture of control system. A set of techniques for collision mapping of the scene is described and impl...
Conference Paper
Full-text available
We present a framework for dynamic quadrupedal locomotion over challenging terrain, where the choice of appropriate footholds is crucial for the success of the behaviour. We build a model of the environment on-line and on-board using an efficient occupancy grid representation. We use Any-time-Repairing A∗ (ARA∗) to search over a tree of possible ac...
Conference Paper
Full-text available
We present a legged motion planning approach for quadrupedal locomotion over challenging terrain. We decompose the problem into body action planning and footstep planning. We use a lattice representation together with a set of defined body movement primitives for computing a body action plan. The lattice representation allows us to plan versatile m...
Conference Paper
Full-text available
This paper describes a novel feature extraction method for laser rangefinder data. Extracted features correspond to real and virtual corners of the scanned scene. The method is based on the Hough Transform (HT) for line extraction, where the intersecting points of these lines are considered as features. This work highlights the use of the HT outsid...
Conference Paper
This paper presents the modeling and study of dynamic behavior of a backhoe machine for tuning of PID controller. The tuning procedure of PID controller is performed, in detailed, for the case of a typical operation, digging a foundation and truck loading. This tuning procedure guarantees the local asymptotic stability in the sense of Lyapunov of o...
Thesis
Full-text available
In this work is developed an architecture which consist of one perception system, one motion planner, two control systems and one localization system that permit the autonomous generation of operations in backhoe machines. This set of systems are presented in two parts that are related with hierarchy levels, i.e. control and localization, and perce...
Conference Paper
Full-text available
This paper presents a motion planning and control system architecture devel- opment for autonomous earthmoving operations in excavating machines such as loading a dump truck. The motion planning system is imitation learning based, which is a general approach for learning motor skills from human demonstra- tion. This scheme of supervised learning is...

Network

Cited By

Projects

Projects (4)
Project
The goal of this project is to achieve reliable locomotion behaviors in semi-structured environment with the humanoid robots HRP-2 and Pyrene. This project relies on model-based feedback control techniques, such as inverse-dynamics and model predictive control.