In this paper, we present an extended mathematical model of the central pattern generator (CPG) in the spinal cord. The proposed CPG model is used as the underlying low-level controller of a humanoid robot to generate various walking patterns. Such biological mechanisms have been demonstrated to be robust in locomotion of animal. Our model is supported by two neurophysiological studies. The first study identified a neural circuitry consisting of a two-layered CPG, in which pattern formation and rhythm generation are produced at different levels. The second study focused on a specific neural model that can generate different patterns, including oscillation. This neural model was employed in the pattern generation layer of our CPG, which enables it to produce different motion patterns-rhythmic as well as non-rhythmic motions. Due to the pattern-formation layer, the CPG is able to produce behaviors related to the dominating rhythm (extension/flexion) and rhythm deletion without rhythm resetting. The proposed multi-layered multi-pattern CPG model (MLMP-CPG) has been deployed in a 3D humanoid robot (NAO) while it performs locomotion tasks. The effectiveness of our model is demonstrated in simulations and through experimental results.

Electronic noses are studied and developed since many years, aiming today to enhance the sensitivity floor, the response time, or characterize new chemical processes. Nowadays, the most performant apparatus are cumbersome, expensive, and not fully dedicated to mobile systems. Most of the researches related to embedded noses on moving applications aim to develop mapping or source detection of toxic gases, enhancing the geometry of the nose and taking into account the air flow perturbations. In this study, we aimed to develop a compact electronic nose dedicated to identify volatile and nonvolatile odors. The aim is at midterm to embed it on a humanoid robot. The purpose is to achieve a compact and embedded electronic nose to provide an added function for robots evolving in their daily environment and able to learn more odors. The final goal is mainly to use it in human-like behavior, mostly to sense food with several conditioning possibilities like volatile or nonvolatile odors. The developed E-Nose has four MOS gas sensors, embedded electronics, and software based on k-nearest neighbors' classification algorithm. Experimental results show success rates up to 98% to differentiate between four fruits juices, namely apple, orange, pineapple, and grenade and to identify rotten eggs from good eggs, grenade perfume, and butane gas, in less than 60 s. An extensible interface enables the E-Nose to learn more odors.

This work is focused on the determination of the thumb and the index finger muscle tensions in a tip pinch task. A biomechanical model of the musculoskeletal system of the thumb and the index finger is developed. Due to the assumptions made in carrying out the biomechanical model, the formulated force analysis problem is indeterminate leading to an infinite number of solutions. Thus, constrained single and multi-objective optimization methodologies are used in order to explore the muscular redundancy and to predict optimal muscle tension distributions. Various models are investigated using the optimization process. The basic criteria to minimize are the sum of the muscle stresses, the sum of individual muscle tensions and the maximum muscle stress. The multi-objective optimization is solved using a Pareto genetic algorithm to obtain non-dominated solutions, defined as the set of optimal distributions of muscle tensions. The results show the advantage of the multi-objective formulation over the single objective one. The obtained solutions are compared to those available in the literature demonstrating the effectiveness of our approach in the analysis of the fingers musculoskeletal systems when predicting muscle tensions.