ERC FUNDED PROJECTS

Actanthrope intends to promote a neuro-robotics perspective to explore original models of anthropomorphic action. The project targets contributions to humanoid robot autonomy (for rescue and service robotics), to advanced human body simulation (for applications in ergonomics), and to a new theory of embodied intelligence (by promoting a motion-based semiotics of the human action). Actions take place in the physical space while they originate in the –robot or human– sensory-motor space. Geometry is the core abstraction that makes the link between these spaces. Considering that the structure of actions inherits from that of the body, the underlying intuition is that actions can be segmented within discrete sub-spaces lying in the entire continuous posture space. Such sub-spaces are viewed as symbols bridging deliberative reasoning and reactive control. Actanthrope argues that geometric approaches to motion segmentation and generation as promising and innovative routes to explore embodied intelligence: - Motion segmentation: what are the sub-manifolds that define the structure of a given action? - Motion generation: among all the solution paths within a given sub-manifold, what is the underlying law that makes the selection? In Robotics these questions are related to the competition between abstract symbol manipulation and physical signal processing. In Computational Neuroscience the questions refer to the quest of motion invariants. The ambition of the project is to promote a dual perspective: exploring the computational foundations of human action to make better robots, while simultaneously doing better robotics to better understand human action. A unique “Anthropomorphic Action Factory” supports the methodology. It aims at attracting to a single lab, researchers with complementary know-how and solid mathematical background. All of them will benefit from unique equipments, while being stimulated by four challenges dealing with locomotion and manipulation actions.

2 500 000 €

Start date: 2014-01-01, End date: 2018-12-31

Parameterized mathematical models have been central to the understanding and design of communication, networking, and radar systems. However, they often lack the ability to model intricate interactions innate in complex systems. On the other hand, data-driven approaches do not need explicit mathematical models for data generation and have a wider applicability at the cost of flexibility. These approaches need labelled data, representing all the facets of the system interaction with the environment. With the aforementioned systems becoming increasingly complex with intricate interactions and operating in dynamic environments, the number of system configurations can be rather large leading to paucity of labelled data. Thus there are emerging networks of systems of critical importance whose cognition is not effectively covered by traditional approaches. AGNOSTIC uses the process of exploration through system probing and exploitation of observed data in an iterative manner drawing upon traditional model-based approaches and data-driven discriminative learning to enhance functionality, performance, and robustness through the notion of active cognition. AGNOSTIC clearly departs from a passive assimilation of data and aims to formalize the exploitation/exploration framework in dynamic environments. The development of this framework in three applications areas is central to AGNOSTIC. The project aims to provide active cognition in radar to learn the environment and other active systems to ensure situational awareness and coexistence; to apply active probing in radio access networks to infer network behaviour towards spectrum sharing and self-configuration; and to learn and adapt to user demand for content distribution in caching networks, drastically improving network efficiency. Although these cognitive systems interact with the environment in very different ways, sufficient abstraction allows cross-fertilization of insights and approaches motivating their joint treatment.

2 499 595 €

Start date: 2017-10-01, End date: 2022-09-30

The goal of the ARS project is the derivation of a unified framework for the autonomous execution of robotic tasks in challenging environments in which accurate performance and safety are of paramount importance. We have chosen surgery as the research scenario because of its importance, its intrinsic challenges, and the presence of three factors that make this project feasible and timely. In fact, we have recently concluded the I-SUR project demonstrating the feasibility of autonomous surgical actions, we have access to the first big data made available to researchers of clinical robotic surgeries, and we will be able to demonstrate the project results on the high performance surgical robot “da Vinci Research Kit”. The impact of autonomous robots on the workforce is a current subject of discussion, but surgical autonomy will be welcome by the medical personnel, e.g. to carry out simple intervention steps, react faster to unexpected events, or monitor the insurgence of fatigue. The framework for autonomous robotic surgery will include five main research objectives. The first will address the analysis of robotic surgery data set to extract action and knowledge models of the intervention. The second objective will focus on planning, which will consist of instantiating the intervention models to a patient specific anatomy. The third objective will address the design of the hybrid controllers for the discrete and continuous parts of the intervention. The fourth research objective will focus on real time reasoning to assess the intervention state and the overall surgical situation. Finally, the last research objective will address the verification, validation and benchmark of the autonomous surgical robotic capabilities. The research results to be achieved by ARS will contribute to paving the way towards enhancing autonomy and operational capabilities of service robots, with the ambitious goal of bridging the gap between robotic and human task execution capability.

2 750 000 €

Start date: 2017-10-01, End date: 2023-09-30

To obtain data-driven dynamic models for simulation, prediction, monitoring, classification or control tasks, in applications e.g. in Industry 4.0 and eHealth, most identification methods ‘solve’ an optimization problem, relying on some nonlinear iterative algorithm. Undeniably, too many heuristics prevail: What do we mean by ‘solved’? Where did the algorithm converge to? Is the model globally optimal, unique and reproducible? To tackle these scientific deficiencies, we design a framework to deal with inexact data. We solve a longstanding open problem of least squares optimality in system identification: for polynomial dynamical models, the optimal model derives from an eigenvalue problem. Hereto, we generalize notions from Algebraic Geometry (multivariate polynomials), Operator Theory (model spaces), System Theory (multidimensional realization) and Numerical Linear Algebra (matrix computations). The first objective is to develop a mathematically rigorous realization approach that maps data onto new mathematical structures (multi-shift invariant projective subspaces). The second objective is to conceive a ‘misfit-latency’ framework to optimally map inexact data to these mathematical structures. We prove this to be a multiparameter eigenvalue problem. We expect breakthroughs in system theoretic characterizations of optimality (covering all existing methods), in the generalization to multiple input-output and multidimensional models and in finding the global optimum in the linear dynamic H2 model reduction problem. The third objective is to implement matrix computation algorithms for the results of the first two objectives, to root sets of multivariate polynomials, to solve multiparameter eigenvalue problems and to isolate only the minimizing roots. We focus on matrix aspects of large scale, sparsity and structure. Deliverables will be publications, software, graduate course material and science outreach initiatives, in line with the PI’s excellent track record

2 485 925 €

Start date: 2021-01-01, End date: 2025-12-31