Pablo Marin-Plaza

Currently, reduced mobility in the upper limbs limits the autonomy of many individuals – including stroke survivors, those with trauma or injuries resulting from accidents, people with neuromuscular diseases, and the elderly experiencing a decline in mobility. Although rehabilitation involving repetitive, adapted exercises is key to motor recovery, intensive therapy often requires continuous access to specialized personnel or complex commercial devices. This entails financial and logistical demands that are not within everyone’s reach. This chapter presents the design and implementation of a portable rehabilitation device specialized for active and passive upper-limb exercises for individuals with reduced mobility. The system is based on a motorized crank that supports the upper limb and guides semicircular shoulder and elbow movements, helping to exercise muscles and improve flexibility. The device incorporates a closed-loop control system that adapts to movements, allowing for speed adjustments and providing “Assist-as-Needed” adaptive control. It prioritizes safety by ensuring muscles are never forced, and it offers two operating modes: active and passive. Furthermore, the system is designed using lightweight, low-cost components, making it suitable for use in laboratories, care homes, physiotherapy centers, and private households. The selected tests involve two opposing exercises and two comprehensive experiments designed to validate the data acquisition, control, and safety systems, thereby demonstrating the device’s viability.

Today’s urban environments are complex, highly congested traffic scenarios that suffer from multiple unsolved problems such as traffic jams and congestion. These problems pose a significant increase in the risks and probability of traffic accidents in modern cities, which have experienced an enormous growth in the number of vehicles. This work introduces a centralized arbitration framework designed for Cooperative Connected Automated Vehicles (CCAVs) to make real-time decisions and resolve conflicts among various driving strategies or behaviors to facilitate resource reservation based on their collaborative actions. Cooperation and arbitration are two of the most important areas of research that seek to provide tools and mechanisms for the optimization and control of traffic flow at critical locations such as intersections and traffic circles. The approach presented, fully implemented on ROS and capable of constructing a software-defined traffic control environment, is able to supervise in a distributed manner how any CCAV operates with the infrastructure, potentially reducing the number of vehicles waiting and harmonizing the traffic flow. The methodology proposed surpasses traditional driver-in-the-loop cooperation by delivering a higher level of automation for collaborative traffic behavior. This approach demonstrably reduces average waiting time by 13% and increases the total utilization of the traffic emplacement by 70% compared to the classic simulated traffic light model. The solution presented was tested on the Carla simulator, with a complete ROS-based vehicle automation solution that provides promising results for CCAV coordination in complex traffic scenarios through a general framework of behavior-based collaboration.

Equipping autonomous agents for dynamic interaction and navigation is a significant challenge in intelligent transportation systems. This study aims to address this by implementing a brain-inspired model for decision making in autonomous vehicles. We employ active inference, a Bayesian approach that models decision-making processes similar to the human brain, focusing on the agent’s preferences and the principle of free energy. This approach is combined with imitation learning to enhance the vehicle’s ability to adapt to new observations and make human-like decisions. The research involved developing a multi-modal self-awareness architecture for autonomous driving systems and testing this model in driving scenarios, including abnormal observations. The results demonstrated the model’s effectiveness in enabling the vehicle to make safe decisions, particularly in unobserved or dynamic environments. The study concludes that the integration of active inference with imitation learning significantly improves the performance of autonomous vehicles, offering a promising direction for future developments in intelligent transportation systems.

Classical imitation learning methods suffer substantially from the learning hierarchical policies when the imitative agent faces an unobserved state by the expert agent. To address these drawbacks, we propose an online active learning through active inference approach that encodes the expert’s demonstrations based on observation-action to improve the learner’s future motion prediction. For this purpose, we provide a switching Dynamic Bayesian Network based on the dynamic interaction between the expert agent and another object in its surrounding as a reference model, which we exploit to initialize an incremental probabilistic learning model. This learning model grows and matures based on the free-energy formulation and message passing of active inference dynamically at discrete and continuous levels in an online active learning phase. In this scheme, generalized states of the learning world are represented as distance-vector, where it is the learner’s observation concerning its interaction with a moving object. Considering the distance vector entail intentions, it enables action prediction evaluation in a prospective sense. We illustrate these points using simulations of driving intelligent agents. The learning agent is trained by using long-term predictions from the generative learning model to reproduce the expert’s motion while learning how to select a suitable action through new experiences. Our results affirm that a Dynamic Bayesian optimal approach provides a principled framework and outperforms conventional reinforcement learning methods. Furthermore, it endorses the general formulation of action prediction as active inference.

This paper proposes an algorithm to model and process streams of LiDAR data under an autonomous vehicle framework. LiDAR is assumed to be an exteroceptive sensor that allows the vehicle to have dynamic 3D scene perception of its surroundings. We employ an encoder-decoder architecture based on 3D-Convolutional layers called 3D Convolution Encoder-Decoder (3D-CED), together with a transfer learning strategy to extract a set of features from point clouds, which are relevant in the context of autonomous driving. The resulting features allow to make inferences of the future point cloud data and detect multiple abstraction level anomalies in controlled scenarios by utilizing a probabilistic switching dynamic model called High Dimensional Markov Jump Particle Filter (HD-MJPF). Moreover, a comparison is provided between piecewise linear, piecewise nonlinear, and nonlinear predictive models for anomaly detection at multiple abstraction levels. Our approach is evaluated with data collected from the LiDAR sensors of the autonomous vehicle while performing certain tasks in a controlled environment.

The expansion of electric vehicles in urban areas has paved the way toward the era of autonomous vehicles, improving the performance in smart cities and upgrading related driving problems. This field of research opens immediate applications in the tourism areas, airports or business centres by greatly improving transport efficiency and reducing repetitive human tasks. This project shows the problems derived from autonomous driving such as vehicle localization, low coverage of 4G/5G and GPS, detection of the road and navigable zones including intersections, detection of static and dynamic obstacles, longitudinal and lateral control and cybersecurity aspects. The approaches proposed in this article are sufficient to solve the operational design of the problems related to autonomous vehicle application in the special locations such as rough environment, high slopes and unstructured terrain without traffic rules.

The advancements in connected and autonomous vehicles in these times demand the availability of tools providing the agents with the capability to be aware and predict their own states and context dynamics. This article presents a novel approach to develop an initial level of collective awareness in a network of intelligent agents. A specific collective self awareness functionality is considered, namely, agent centered detection of abnormal situations present in the environment around any agent in the network. Moreover, the agent should be capable of analyzing how such abnormalities can influence the future actions of each agent. Data driven dynamic Bayesian network (DBN) models learned from time series of sensory data recorded during the realization of tasks (agent network experiences) are here used for abnormality detection and prediction. A set of DBNs, each related to an agent, is used to allow the agents in the network to each synchronously aware possible abnormalities occurring when available models are used on a new instance of the task for which DBNs have been learned. A growing neural gas (GNG) algorithm is used to learn the node variables and conditional probabilities linking nodes in the DBN models; a Markov jump particle filter (MJPF) is employed for state estimation and abnormality detection in each agent using learned DBNs as filter parameters. Performance metrics are discussed to asses the algorithms reliability and accuracy. The impact is also evaluated by the communication channel used by the network to share the data sensed in a distributed way by each agent of the network. The IEEE 802.11p protocol standard has been considered for communication among agents. Real data sets are also used acquired by autonomous vehicles performing different tasks in a controlled environment.

This work presents a novel platform for autonomous vehicle technologies research for the insurance sector. The platform has been collaboratively developed by the insurance company MAPFRE-CESVIMAP, Universidad Carlos III de Madrid and INSIA of the Universidad Politécnica de Madrid. The high-level architecture and several autonomous vehicle technologies developed using the framework of this collaboration are introduced and described in this work. Computer vision technologies for environment perception, V2X communication capabilities, enhanced localization, human–machine interaction and self awareness are among the technologies which have been developed and tested. Some use cases that validate the technologies presented in the platform are also presented; these use cases include public demonstrations, tests of the technologies and international competitions for self-driving technologies.