Search

Despite a significant proportion of the Earth being covered in water, exploration of what lies below has been limited due to the challenges and difficulties inherent in the process. Current state of the art robots such as Remotely Operated Vehicles (ROVs) and Autonomous Underwater Vehicles (AUVs) are bulky, rigid and unable to conform to their environment. Soft robotics offers solutions to this issue. Fluid-actuated eversion or growing robots, in particular, are a good example. While current eversion robots have found many applications on land, their inherent properties make them particularly well suited to underwater environments. An important factor when considering underwater eversion robots is the establishment of a suitable steering mechanism that can enable the robot to change direction as required. This project proposes a design for an eversion robot that is capable of steering while underwater, through the use of bending pouches, a design commonly seen in the literature on land-based eversion robots. These bending pouches contract to enable directional change. Similar to their land-based counterparts, the underwater eversion robot uses the same fluid in the medium it operates in to achieve extension and bending but also to additionally aid in neutral buoyancy. The actuation method of bending pouches meant that robots needed to fully extend before steering was possible. Three robots, with the same design and dimensions were constructed from polyethylene tubes and tested. Our research shows that although the soft eversion robot design in this paper was not capable of consistently generating the same amounts of bending for the inflation volume, it still achieved suitable bending at a range of inflation volumes and was observed to bend to a maximum angle of 68 degrees at 2000 ml, which is in line with the bending angles reported for land-based eversion robots in the literature.

Class incremental learning consists in training discriminative models to classify an increasing number of classes over time. However, doing so using only the newly added class data leads to the known problem of catastrophic forgetting of the previous classes. Recently, dynamic deep learning architectures have been shown to exhibit a better stability-plasticity trade-off by dynamically adding new feature extractors to the model in order to learn new classes followed by a compression step to scale the model back to its original size, thus avoiding a growing number of parameters. In this context, we propose a new algorithm that enhances the compression of previous class knowledge by cutting and mixing patches of previous class samples with the new images during compression using our Rehearsal-CutMix method. We show that this new data augmentation reduces catastrophic forgetting by specifically targeting past class information and improving its compression. Extensive experiments performed on the CIFAR and ImageNet datasets under diverse incremental learning evaluation protocols demonstrate that our approach consistently outperforms the state-of-the-art . The code will be made available upon publication of our work.

Underwater Unmanned Vehicles (UUVs) have to constantly compensate for the external disturbing forces acting on their body. Adaptive Control theory is commonly used there to grant the control law some flexibility in its response to process variation. Today, learning-based (LB) adaptive methods are leading the field where model-based control structures are combined with deep model-free learning algorithms. This work proposes experiments and metrics to empirically study the stability of such a controller. We perform this stability analysis on a LB adaptive control system whose adaptive parameters are determined using a Cross-Entropy Deep Learning method.

Autonomous vessels have emerged as a prominent and accepted solution, particularly in the naval defence sector. However, achieving full autonomy for marine vessels demands the development of robust and reliable control and guidance systems that can handle various encounters with manned and unmanned vessels while operating effectively under diverse weather and sea conditions. A significant challenge in this pursuit is ensuring the autonomous vessels' compliance with the International Regulations for Preventing Collisions at Sea (COLREGs). These regulations present a formidable hurdle for the human-level understanding by autonomous systems as they were originally designed from common navigation practices created since the mid-19th century. Their ambiguous language assumes experienced sailors' interpretation and execution, and therefore demands a high-level (cognitive) understanding of language and agent intentions. These capabilities surpass the current state-of-the-art in intelligent systems. This position paper highlights the critical requirements for a trustworthy control and guidance system, exploring the complexity of adapting COLREGs for safe vessel-on-vessel encounters considering autonomous maritime technology competing and/or cooperating with manned vessels.

Deep Reinforcement Learning (DRL) techniques have received significant attention in control and decision-making algorithms. Most applications involve complex decision-making systems, justified by the algorithms' computational power and cost. While model-based versions are emerging, model-free DRL approaches are intriguing for their independence from models, yet they remain relatively less explored in terms of performance, particularly in applied control. This study conducts a thorough performance analysis comparing the data-driven DRL paradigm with a classical state feedback controller, both designed based on the same cost (reward) function of the linear quadratic regulator (LQR) problem. Twelve additional performance criteria are introduced to assess the controllers' performance, independent of the LQR problem for which they are designed. Two Deep Deterministic Policy Gradient (DDPG)-based controllers are developed, leveraging DDPG's widespread reputation. These controllers are aimed at addressing a challenging setpoint tracking problem in a Non-Minimum Phase (NMP) system. The performance and robustness of the controllers are assessed in the presence of operational challenges, including disturbance, noise, initial conditions, and model uncertainties. The findings suggest that the DDPG controller demonstrates promising behavior under rigorous test conditions. Nevertheless, further improvements are necessary for the DDPG controller to outperform classical methods in all criteria. While DRL algorithms may excel in complex environments owing to the flexibility in the reward function definition, this paper offers practical insights and a comparison framework specifically designed to evaluate these algorithms within the context of control engineering.

Learning-based adaptive control methods hold the premise of enabling autonomous agents to reduce the effect of process variations with minimal human intervention. However, its application to autonomous underwater vehicles (AUVs) has so far been restricted due to 1) unknown dynamics under the form of sea current disturbance that we can not model properly nor measure due to limited sensor capability and 2) the nonlinearity of AUVs tasks where the controller response at some operating points must be overly conservative in order to satisfy the specification at other operating points. Deep Reinforcement Learning (DRL) can alleviates these limitations by training general-purpose neural network policies, but applications of DRL algorithms to AUVs have been restricted to simulated environments, due to their inherent high sample complexity and distribution shift problem. This paper presents a novel approach, merging the Maximum Entropy Deep Reinforcement Learning framework with a classic model-based control architecture, to formulate an adaptive controller. Within this framework, we introduce a Sim-to-Real transfer strategy comprising the following components: a bio-inspired experience replay mechanism, an enhanced domain randomisation technique, and an evaluation protocol executed on a physical platform. Our experimental assessments demonstrate that this method effectively learns proficient policies from suboptimal simulated models of the AUV, resulting in control performance 3 times higher when transferred to a real-world vehicle, compared to its model-based nonadaptive but optimal counterpart.

The bouncing neutron is one of the rare system where gravity can be studied in a quantum framework. To this end it is crucial to be able to select some specific gravitational quantum state (GQS). The GRANIT apparatus is the first physics experiment connected to a superthermal helium UCN source. We report on the methods developed for this instrument showing how specific GQS can be favored using a step between mirrors and an absorbing slit. We explore the increase of GQS separation efficiency by increasing the absorber roughness amplitude, and find it is feasible but requires a high adjustment precision. We also quantify the transmission of the absorbing slit leading to a measurement of the spatial extension of the neutron vertical wave function $z_0 = \hbar^{2/3}\left(2m^2g\right)^{-1/3} = 5.9\pm0.3\,\mu$m., Comment: 5 pages, 3 figures

Navigation problems under unknown varying conditions are among the most important and well-studied problems in the control field. Classic model-based adaptive control methods can be applied only when a convenient model of the plant or environment is provided. Recent model-free adaptive control methods aim at removing this dependency by learning the physical characteristics of the plant and/or process directly from sensor feedback. Although there have been prior attempts at improving these techniques, it remains an open question as to whether it is possible to cope with real-world uncertainties in a control system that is fully based on either paradigm. We propose a conceptually simple learning-based approach composed of a full state feedback controller, tuned robustly by a deep reinforcement learning framework based on the Soft Actor-Critic algorithm. We compare it, in realistic simulations, to a model-free controller that uses the same deep reinforcement learning framework for the control of a micro aerial vehicle under wind gust. The results indicate the great potential of learning-based adaptive control methods in modern dynamical systems.

We report on the development of a 10B conversion layer optimized for ultra-cold neutron detection with silicon detectors. The efficiency of this layer is high and roughly uniform over a large ultra-cold neutron velocity range. The designed titanium-boron-nickel multilayer film was deposited on silicon using a microwave plasma-assisted co-sputtering method (first, for test purpose, on silicon wafers, then directly on the surface of a CCD sensor). The obtained sensor was then tested using both cold and ultra-cold neutrons., Comment: 12 pages

A time-of-flight device was developed to characterize wavelength distribution and uniformity of a cold neutron beam. This device is very compact -- the distance of flight is $60$ cm -- but achieves very high resolution -- the intrinsic resolution $\Delta \lambda/\lambda=2.4\cdot 10^{-3}$ at $\lambda=0.89$ nm. The time-of-flight device is composed of a fixed slit, a disk rotating up to $216$ Hz and a neutron detector with a thin spherical conversion layer with the chopper slit in its focus. The device accepts the complete angular divergence of the initial neutron beam. The efficiency of neutron detection is constant over the detector area. Systematic corrections caused by neutron scattering in air are minimized due to the reduction of the time-of-flight length. Measurements have been performed on the beamline of the GRANIT experiment at ILL (part of the H172 beamline) on level C, and the first order diffraction peak of the crystal monochromator used for the GRANIT beamline was found to be at $\lambda=0.8961(11)$ nm, and having a width of $\sigma=0.0213(13)$ nm.

The GRANIT facility is a follow-up project, which is motivated by the recent discovery of gravitational quantum states of ultracold neutrons. The goal of the project is to approach the ultimate accuracy in measuring parameters of such quantum states and also to apply this phenomenon and related experimental techniques to a broad range of applications in particle physics as well as in surface and nanoscience studies. We overview the current status of this facility, the recent test measurements and the nearest prospects., Comment: 11 pages, 20 figures, Proceedings of the GRANIT-2014 WORKSHOP "Quantum gravitational spectroscopy with ultracold systems"(Les Houches)