Research Topics

Key research directions that our lab pursues.

Key research topics along with representative publications are listed by topics below -- full publication list here. Please see an introduction to our lab in the highlighted video below. Full collection of SLMC research videos on the dedicated YouTube Playlist.


Variable Impedance Modulation: An Algorithmic Approach to Optimality

The ability to modulate Impedance (stiffness and damping) is a powerful tool to enable safe and efficient actuation. Our group has world leading expertise in the algorithmic optimisation of spatiotemporal impedance modulation for a range of hardware platforms and varying motion objectives.

  • Franco Angelini, Guiyan Xin, Wouter Wolfslag, Carlo Tiseo, Michael Mistry, Manolo Garabini, Antonio Bicchi and Sethu Vijayakumar, Online Optimal Impedance Planning for Legged Robots, Proc. IEEE International Conf. on Intelligent Robots and Systems (IROS 2019), Macau, China (2019). [pdf] [video]

  • David Braun, Florian Petit, Felix Huber, Sami Haddadin, Patrick van der Smagt, Alin Albu-Schaeffer and Sethu Vijayakumar, Robots Driven by Compliant Actuators: Optimal Control under Actuation Constraints, IEEE Transactions on Robotics (IEEE T-RO), 29(5), pp. 1085-1101 (2013). [pdf] [video] [2013 IEEE Transactions on Robotics Best Paper award]

  • Jun Nakanishi, Konrad Rawlik and Sethu Vijayakumar, Stiffness and Temporal Optimization in Periodic Movements: An Optimal Control Approach, Proc. IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS 2011, San Francisco (2011). [pdf] [video]


Topology-based Representations for Efficient Planning and Adaptation

Our group has developed methods to combine and exploit different toplogical representations for motion synthesis, with specific emphasis on generalization of motion to novel situations. We have focused on problems where direct path finding in joint configuration space is extremely hard whereas exploiting a representation with different topology can find optimal trajectories with efficient, local methods.

  • Vladimir Ivan, Dmitry Zarubin, Marc Toussaint and Sethu Vijayakumar, Topology-based Representations for Motion Planning and Generalisation in Dynamic Environments with Interactions, International Journal of Robotics Research (IJRR), vol. 32, pp. 1151-1163 (2013). [pdf] [DOI] [video]

  • Masashi Sugiyama, Hirotaka Hachiya, Christopher Towell and Sethu Vijayakumar, Geodesic Gaussian kernels for value function approximation, Autonomous Robots, Vol. 25, pp. 287-304 (2008). [pdf][DOI]

  • Sebastian Bitzer, Matthew Howard and Sethu Vijayakumar, Using Dimensionality Reduction to Exploit Constraints in Reinforcement Learning, Proc. IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS 2010), Taiwan (2010). [pdf][video]


Adaptive Real-time Motion Planning for Dynamic Environments

A keystone of our research approach is adaptability to dynamic environments. We have developed several online adaptation methods that can plan efficient trajectories in response to changing environments and model mismatches -- by incorporating online learning methods as well as offline computations with efficient indexing.

  • Yiming Yang, Wolfgang Merkt, Vladimir Ivan, and Sethu Vijayakumar, Planning in Time-Configuration Space for Efficient Pick-and-Place in Non-Static Environments with Temporal Constraints, Proc. IEEE-RAS 18th Intl. Conf. on Humanoid Robots (Humanoids 2018), Beijing, China (2018) [pdf] [video]

  • Andreea Radulescu, Jun Nakanishi, David Braun and Sethu Vijayakumar, Optimal Control of Variable Stiffness Policies: Dealing with Switching Dynamics and Model Mismatch, J.-P. Laumond et al. (eds.), Geometric and Numerical Foundations of Movements, Springer Tracts in Advanced Robotics 117, pp. 393-419 (2017). [DOI] [pdf]

  • Yiming Yang, Vladimir Ivan, Zhibin Li, Maurice Fallon and Sethu Vijayakumar, iDRM: Humanoid Motion Planning with Realtime End-Pose Selection in Complex Environments, Proc. 16th IEEE-RAS International Conference on Humanoid Robots, Cancun, Mexico (2016) [pdf][video]


Collaborative Contact-Aware Manipulation (Multi-agent, Human-Robot)

Humans are able to collaborate intuitively and effortlessly. To equip robots to do the same, we are tackling several problems ranging from human intention detection, dyadic collaborations that are partner aware to  realtime planning of hybrid trajectory optimisation problems (contact sequence, grasp hold changes, ergonomic trajectories) that can make human-robot collaborations more natural.

  • Theodoros Stouraitis, Lei Yan, Joao Moura, Michael Gienger and Sethu Vijayakumar, Multi-mode Trajectory Optimization for Impact-aware Manipulation, Proc. IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS 2020), Las Vegas, USA (2020). [pdf] [video] [digest]

  • Theodoros Stouratis, Iordanis Chatzinikolaidis, Michael Gienger and Sethu Vijayakumar, Dyadic collaborative Manipulation through Hybrid Trajectory Optimization, Proc. Conference on Robot Learning (CoRL 2018), Madrid (2018). [pdf][video] [CoRL 2018 Best Paper Award finalist]

  • Lei Yan, Yiming Yang, Wenfu Xu, and Sethu Vijayakumar, Dual-arm Coordinated Motion Planning and Compliance Control for Capturing Moving Objects with Large Momentum, Proc. IEEE Intl. Conf. on Intelligent Robotics (IROS 2018), Madrid (2018). [pdf] [video]


Constraint Consistent Learning from Demonstrations

Learning from demonstrations (LfD) is a powerful way of bootstrapping solutions. There are challenges in making most efficient use of the demonstrations, especially when it comes from varying constrained observations. Our research aims to create a theory of principled, constraint consistent learning that can be exploited over various domains and kinematic/kino-dynamic tasks.

  • Joao Moura, Vladimir Ivan, Mustafa Suphi Erden and Sethu Vijayakumar, Equivalence of the Projected Forward Dynamics and the Dynamically Consistent Inverse Solution, Proc. Robotics: Science and Systems (RSS 2019), Frieberg, Germany (2019). [pdf][poster] [RSS 2019 Best Paper Award Finalist]

  • Leopoldo Armesto, Joao Moura, Vladimir Ivan, Mustafa Suphi Erden, Antonio Sala and Sethu Vijayakumar, Constraint-aware Learning of Policies by Demonstration, International Journal of Robotics Research (IJRR), Vol. 37(13-14), pp. 1673-1689 (2018). [pdf][video]

  • Matthew Howard, David J. Braun, Sethu Vijayakumar, Transferring Human Impedance Behaviour to Heterogeneous Variable Impedance Actuators, IEEE Transactions on Robotics, 29(4), pp. 847-862 (2013). [pdf]


Data-driven Online Incremental Learning for Dynamics Adaptation 

Advances in Machine Learning (ML) has given us very powerful tools to learn dynamic and complex models from observed data. Our research focuses on developing non-parametric ML techniques that cater to the unique demands of robot learning -- namely, incremental, online, memory efficient and real-time methods that have reasonable guarantees against catastrophic failure (graceful degradation).

  • Djordje Mitrovic, Stefan Klanke and Sethu Vijayakumar, Learning Impedance Control of Antagonistic Systems based on Stochastic Optimisation Principles, International Journal of Robotic Research (IJRR), Vol. 30, No. 5, pp. 556-573 (2011). [pdf] [DOI]

  • Kian Ming Chai, Stefan Klanke, Chris Williams and Sethu Vijayakumar, Multi-task Gaussian Process Learning of Robot Inverse Dynamics, Proc. Advances in Neural Information Processing Systems (NIPS '08), Vancouver, Canada (2008).[pdf]

  • Sethu Vijayakumar and Stefan Schaal, LWPR : An O(n) Algorithm for Incremental Real Time Learning in High Dimensional Space, Proc. of Seventeenth International Conference on Machine Learning (ICML2000) California, USA, pp.1079-1086 (2000) [ pdf]


Loco-manipulation in High Dimensions [Valkyrie, ANYmal, ASIMO]

Scaling whole-body motion planning in a consistent, coherent and holistic manner gets tougher in high-dimensional robotic platforms. Our group conducts research on some of the world's most complex humanoid and legged platforms, where we test whole-body, multi contact motion planning methods for real-world scalability.

  • Henrique Ferrolho, Wolfgang Xaver Merkt, Vladimir Ivan, Wouter Wolfslag and Sethu Vijayakumar, Optimizing Dynamic Trajectories for Robustness to Disturbances using Polytopic Projections, Proc. IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS 2020), Las Vegas, USA (2020). [pdf] [video] [digest]

  • Yiming Yang, Wolfgang Merkt, Henrique Ferrolho, Vladimir Ivan and Sethu Vijayakumar, Efficient Humanoid Motion Planning on Uneven Terrain Using Paired Forward-Inverse Dynamic Reachability Maps, IEEE Robotics and Automation Letters, vol. 2(4), pp. 2279-2286 (2017) [pdf][DOI][video] (also presented at: IEEE Intl. Conf. on Intelligent Robots and Systems (IROS'17), Vancouver, Canada (2017)).

  • Matthew Howard, Stefan Klanke, Michael Gienger, Christian Goerick and Sethu Vijayakumar, A Novel Method for Learning Policies from Variable Constraint Data, Autonomous Robots, vol. 27, pp. 105-121 (2009). [pdf] [ASIMO car wash video]


Bipedal Locomotion: In Human and Robots, Wheels and Legs

Bipedal locomotion has held a special fascination in robotics research. It is complex, tests the reactiveness of methods and requires efficient  planning and control strategies. Research in this domain has the promise to unlock some of the deepest mysteries and idiosyncracies seen in human locomotion while also being able to benefit from abstraction/strategies derived from human motor control.

  • Songyang Xin and Sethu Vijayakumar, Online Dynamic Motion Planning and Control for Wheeled Biped Robots, Proc. IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS 2020), Las Vegas, USA (2020). [pdf] [video] [digest]

  • Henrique Ferrolho,Wolfgang Merkt, Yiming Yang, Vladimir Ivan, and Sethu Vijayakumar, Whole-Body End-Pose Planning for Legged Robots on Inclined Support Surfaces in Complex Environments, Proc. IEEE-RAS 18th International Conference on Humanoid Robots (Humanoids), Beijing, China (2018) [pdf] [video]

  • Christopher McGreavy, Kai Yuan, Daniel F. N. Gordon, Kang Tan, Wouter Wolfslag, Sethu Vijayakumar and Zhibin Li, Unified Push Recovery Fundamentals: Inspiration from Human Study, Proc. IEEE International Conference on Robotics and Automation (ICRA 2020), Paris, France (2020). [pdf]


Multi-contact Whole Body Motion Planning and Control

Humans and Biological Systems are very good at not just avoiding obstacles and contacts but exploiting them to their advantage -- for stability or realising difficult, complicated maneuvers. Our research aims to find efficient motion planning and control metholdologies that can do the same for highly dynamic movements in robots -- quadrupeds, bipeds and other loco-manipulation platforms.

  • Wouter Wolfslag, Christopher McGreavy, Guiyang Xin, Carlo Tiseo, Sethu Vijayakumar and Zhibin Li, Optimisation of Body-ground Contact for Augmenting Whole-Body Loco-manipulation of Quadruped Robots, Proc. IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS 2020), Las Vegas, USA (2020). [pdf] [video] [digest]

  • Carlos Mastalli, Rohan Budhiraja, Wolfgang Xaver Merkt, Guilhem Saurel, Bilal Hammoud, Maximilien Naveau, Justin Carpentier, Ludovic Righetti, Sethu Vijayakumar and Nicolas Mansard, Crocoddyl: An Efficient and Versatile Framework for Multi-Contact Optimal Control, Proc. IEEE International Conference on Robotics and Automation (ICRA 2020), Paris, France (2020). [pdf] [video]

  • Jun Nakanishi, Andreea Radulescu and Sethu Vijayakumar, Spatiotemporal Optimisation of Multi-phase Movements: Dealing with Contacts and Switching Dynamics, Proc. IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS 2013), Tokyo (2013). [pdf] [video]


Novel Hardware Designs and Control for Variable Impedance Actuation

Our group has explored co-design of variable impedance (stiffness and damping) actuation mechanisms that can exploit their spatiotemporal modulation to create robust and efficient motion. Coupling this with our impedance optimisation framework leads to a new class of mechanisms with unprecedented capabilities.

  • Alexander Enoch and Sethu Vijayakumar, Rapid Manufacture of Novel Variable Impedance Robots, ASME Journal of Mechanisms and Robotics, (8)1, pp. 1-11, (2016). [pdf] [DOI]

  • David Braun, Andrius Sutas and Sethu Vijayakumar, Self-tuning Bistable Parametric Feedback Oscillator: Near-optimal Amplitude Maximization without Model Information, Physics Review E, vol. 95(1), 12201 (2017). [pdf] [DOI]

  • Andreaa Radulescu, Matthew Howard, David Braun and Sethu Vijayakumar, Exploiting Variable Physical Damping in Rapid Movement Tasks, Proc. 2012 IEEE ASME International Conference on Advanced Intelligent Mechatronics, Taiwan (2012). [pdf] [video] [AIM 2012 Best Student Paper Award Finalist]


Perception and Muti-sensory Scene Understanding

Our group has worked on novel techniques for sensing, tracking and localisation with the objective of enabling real-time feedback for our planning and control methodologies. Specific attention is paid to realistic scenarios that involve dynamic objects that are either involved as navigation markers or being actively interacted with by our robotic platforms.

  • Karl Pauwels, Vladimir Ivan, Eduardo Ros and Sethu Vijayakumar, Real-time object pose recognition and tracking with an imprecisely calibrated moving RGB-D camera, Proc. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2014), Chicago (2014). [pdf][video]

  • Timothy Hospedales and Sethu Vijayakumar, Bayesian Structure Inference for Multisensory Scene Understanding, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 30, no. 12, pp. 2140-2157 (2008). [pdf][DOI]

  • Heiko Hoffmann, Georgios Petkos, Sebastian Bitzer and Sethu Vijayakumar, Sensor Assisted Adaptive Motor Control under continuously varying Context, Proc. Fourth International Conference on Informatics in Control, Automation and Robotics (ICINCO'07), Angers, France (2007). [pdf]


Probabilistic Inference based Stochastic Optimal Control and Reinforcement Learning

Our group provided one of the earliest analogies between classical Reinforcement Learning and Optimal Control algorithms in terms of  probablisitic approximate inference. This enabled us to deal with inherent actuation and sensing uncertainty in a systematic manner while being able to bring notions such as risk and weighted rewards to bear upon motor control problems.

  • Konrad Rawlik, Marc Toussaint and Sethu Vijayakumar, On Stochastic Optimal Control and Reinforcement Learning by Approximate Inference, Proc. Robotics: Science and Systems (R:SS 2012), Sydney, Australia(2012). [pdf] [R:SS 2012 Best Paper Award Runner-up]

  • Djordje Mitrovic, Stefan Klanke, Rieko Osu, Mitsuo Kawato and Sethu Vijayakumar, A Computational Model of Limb Impedance Control based on Principles of Internal Model Uncertainty, PLoS ONE, Vol. 5, No. 10 (2010). [pdf] [DOI]

  • Konrad Rawlik, Marc Toussaint and Sethu Vijayakumar, An Approximate Inference Approach to Temporal Optimization in Optimal Control, Proc. Advances in Neural Information Processing Systems (NIPS '10), Vancouver, Canada(2010). [pdf] [poster]


Prosthetics & Exoskeletons: Upper Limb

We have worked on several classes of upper limb prosthetics and looked at most efficient ways of providing effective feedback, most intuitive control interfaces as well as discovered new ways of combining multi-modal bio-signals to improve intuitive control of prosthetics. The main focus has been to reduce cognitive load and improve robustness of operations.

  • Agamemnon Krasoulis, Iris Kyranou, Mustapha Suphi Erden, Kianoush Nazarpour and Sethu Vijayakumar, Improved Prosthetic Hand Control with Concurrent Use of Myoelectric and Inertial Measurements, Journal of NeuroEngineering and Rehabilitation (JNER), 14:71 (2017). [pdf][video]

  • Ian Saunders and Sethu Vijayakumar, The Role of Feed-Forward and Feedback Processes for Closed-Loop Prosthesis Control, Journal of Neuroengineering and Rehabilitation (JNER),8:60 (2011). [pdf] [DOI]

  • Agamemnon Krasoulis, Sethu Vijayakumar and Kianoush Nazarpour, Multi-Grip Classification-Based Prosthesis Control With Two EMG-IMU Sensors, IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 28:2, pp. 508-518 (2020) [pdf] [DOI]


Prosthetics & Exoskeletons: Lower Limb

Our group researches lower limb exoskeleton control and gait analysis. Exploiting our world leading gait lab facilities (split treadmill, force plates, vicon motion tracking, EMG sensing, interactive visual feedback, calorimetry, FES), we have experimented with several novel ankle, pelvis and whole lower limb exoskeleton and prosthetic devices including design and evaluations.

  • Daniel Gordon, Graham Henderson and Sethu Vijayakumar, Effectively quantifying the performance of lower-limb exoskeletons over a range of walking conditions, Frontiers in Robotics and AI, vol. 5, pp. 61-77 (2018) [pdf] [DOI] [video]

  • Francois Heremans, Sethu Vijayakumar, Mohamed Bouri, Bruno Dehez and Renaud Ronsse, Bio-inspired design and validation of the Efficient Lockable Spring Ankle (ELSA) prosthesis, Proc. 16th IEEE Intl. Conf. on Rehabilitation Robotics(ICORR 2019), Toronto, Canada (2019). [pdf][video]

  • Hsiu-Chin Lin, Matthew Howard and Sethu Vijayakumar, A Novel Approach for Representing and Generalising Periodic Gaits, Robotica, 32:08, pp. 1225-1244 (2014). [pdf] [DOI]


Shared and Punctuated Autonomy for Remote (Distal) Operations

The ability to seamleslys shift between teleoperation and full autonomy is ever becoming more important with hybrid robotic platforms and capabilities. We study key design issues as well as multi-objective optimisation paradigms  that look at how to blend control inputs while taking into account real-time sensory feedback.

  • Wolfgang Merkt, Yiming Yang, Theodoros Stouraitis, Christopher Mower, Maurice Fallon and Sethu Vijayakumar, Robust Shared Autonomy for Mobile Manipulation with Continuous Scene Monitoring, Proc. 13th IEEE Conference on Automation Science and Engineering, Xian, China (2017). [pdf] [video] [First Prize at Robots for Resilient Infrastructure Challenge 2017, Leeds, UK]

  • Chris Mower, Joao Moura, Aled Davies and Sethu Vijayakumar, Modulating Human Input for Shared Autonomy in Dynamic Environments, Proc. 28th IEEE Intl. Conf. on Robot and Human Interactive Comm.(ROMAN 2019), New Delhi, India (2019). [pdf]

  • Vladimir Ivan, Arnau Garriga-Casanovas, Pouyan Khalili, Wolfgang Merkt, Frederic Cegla and Sethu Vijayakumar, Autonomous Non-destructive Remote Robotic Inspection of Offshore Assets, Proc. Offshore Technology Conference (OTC 2020), Houston, USA (2020). [pdf] [presentation]


Tactile Sensing: In Humans and Robots

Tactile sensing is an important yet not widely explored modality in robotic close contact manipulation. Our work ranges from effective use of tactile sensory data for object discrimination, affordance and contact estimation to study of temporal processing in human tactile feedback.

  • Hannes Saal, Jo-Anne-Ting and Sethu Vijayakumar, Active Estimation of Object Dynamics Parameters with Tactile Sensors, Proc. IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS 2010), Taiwan (2010). [pdf][video]

  • Hannes Saal, Sethu Vijayakumar and Roland Johansson, Information about Complex Fingertip Parameters in Individual Human Tactile Afferent neurons, The Journal of Neuroscience, 29(25):8022-8031, (2009). [pdf]

  • Peter Sandilands, Vladimir Ivan, Taku Komura and Sethu Vijayakumar, Dexterous Reaching, Grasp Transfer and Planning Using Electrostatic Representations, Proc. 2013 IEEE-RAS International Conference on Humanoid Robots, Atlanta, USA (2013). [pdf]


Modelling Human Sensorimotor Control: Generative Model Approaches

We model the multi-modal processing and adaptation in human sensorimotor control using techniques from Bayesian inference. The notion of priors, adaptation and temporal processing from a perspective of optimality os explored through novel experimental design and modelling.

  • Luigi Acerbi, Sethu Vijayakumar and Daniel Wolpert, On the Origins of Suboptimality in Human Probabilistic Inference, PLoS Computational Biology, 10(6): e1003661. doi:10.1371/journal.pcbi.1003661 (2014). [pdf][DOI]

  • Luigi Acerbi, Daniel Wolpert and Sethu Vijayakumar, Internal Representations of Temporal Statistics and Feedback Calibrate Motor-Sensory Interval Timing, PLoS Computational Biology, 8(11): e1002771, doi:10.1371/journal.pcbi.1002771 (2012). [pdf] [DOI]

  • Sethu Vijayakumar, Timothy Hospedales and Adrian Haith, Generative Probabilistic Modeling: Understanding Causal Sensorimotor Integration, In Trommershauser, Kording & Landy (Eds), Sensory Cue Integration, pp. 63-81, Oxford University Press (2011) [OUP] [preprint


Open Source Software packages and Related Works: Control, Planning and Learning 

There are several open source software packages and data sets resulting from our research. We have open sourced and maintained a subset of these that can be downloaded from the links along with illustrative working examples.

  • Vladimir Ivan, Yiming Yang, Wolfgang Merkt, Michael P. Camilleri, Sethu Vijayakumar, EXOTica: An Extensible Optimization Toolset for Prototyping and Benchmarking Motion Planning and Control, In: Koubaa A. (eds) Robot Operating System (ROS). Studies in Computational Intelligence, Springer, vol. 778, pp. 211-240 (2019) [preprint pdf] [software repository]

  • Carlos Mastalli, Rohan Budhiraja, Wolfgang Xaver Merkt, Guilhem Saurel, Bilal Hammoud, Maximilien Naveau, Justin Carpentier, Ludovic Righetti, Sethu Vijayakumar and Nicolas Mansard, Crocoddyl: An Efficient and Versatile Framework for Multi-Contact Optimal Control, Proc. IEEE International Conference on Robotics and Automation (ICRA 2020), Paris, France (2020). [pdf] [video]

  • Stefan Klanke, Sethu Vijayakumar and Stefan Schaal, A Library for Locally Weighted Projection Regression, Journal of Machine Learning Research (JMLR), vol. 9, pp. 623--626 (2008). [pdf][software page]