Until few years ago, #ArtificialIntelligence (#AI) was similar to nuclear fusion in unfulfilled promise. It had been around a long time but had not reached the spectacular heights foreseen in its initial stages. However now, Artificial intelligence (AI) is no longer the future. It's realizing its potential in achieving man-like capabilities, so it's the right time to ask: How can business leaders adapt AI to take advantage of the specific strengths of man and machine? AI is swiftly becoming the foundational technology in areas as diverse as self-driving cars, financial trading, connected houses etc. Self-learning algorithms are now routinely embedded in mobile and online services.

The development of accurate constitutive models for materials that undergo path-dependent processes continues to be a complex challenge in computational solid mechanics. Challenges arise both in considering the appropriate model assumptions and from the viewpoint of data availability, verification, and validation. Recently, data-driven modeling approaches have been proposed that aim to establish stress-evolution laws that avoid user-chosen functional forms by relying on machine learning representations and algorithms. However, these approaches not only require a significant amount of data but also need data that probes the full stress space with a variety of complex loading paths. Furthermore, they rarely enforce all necessary thermodynamic principles as hard constraints. Hence, they are in particular not suitable for low-data or limited-data regimes, where the first arises from the cost of obtaining the data and the latter from the experimental limitations of obtaining labeled data, which is commonly the case in engineering applications. In this work, we discuss a hybrid framework that can work on a variable amount of data by relying on the modularity of the elastoplasticity formulation where each component of the model can be chosen to be either a classical phenomenological or a data-driven model depending on the amount of available information and the complexity of the response. The method is tested on synthetic uniaxial data coming from simulations as well as cyclic experimental data for structural materials. The discovered material models are found to not only interpolate well but also allow for accurate extrapolation in a thermodynamically consistent manner far outside the domain of the training data. Training aspects and details of the implementation of these models into Finite Element simulations are discussed and analyzed.

Nowadays, many AI applications utilizing resource-constrained edge devices (e.g., small mobile robots, tiny IoT devices, etc.) require Convolutional Neural Network (CNN) inference on a distributed system at the edge due to limited resources of a single edge device to accommodate and execute a large CNN. There are four main partitioning strategies that can be utilized to partition a large CNN model and perform distributed CNN inference on multiple devices at the edge. However, to the best of our knowledge, no research has been conducted to investigate how these four partitioning strategies affect the energy consumption per edge device. Such an investigation is important because it will reveal the potential of these partitioning strategies to be used effectively for reduction of the per-device energy consumption when a large CNN model is deployed for distributed inference at the edge. Therefore, in this paper, we investigate and compare the per-device energy consumption of CNN model inference at the edge on a distributed system when the four partitioning strategies are utilized. The goal of our investigation and comparison is to find out which partitioning strategies (and under what conditions) have the highest potential to decrease the energy consumption per edge device when CNN inference is performed at the edge on a distributed system.

Pre-training on time series poses a unique challenge due to the potential mismatch between pre-training and target domains, such as shifts in temporal dynamics, fast-evolving trends, and long-range and short-cyclic effects, which can lead to poor downstream performance. While domain adaptation methods can mitigate these shifts, most methods need examples directly from the target domain, making them suboptimal for pre-training. To address this challenge, methods need to accommodate target domains with different temporal dynamics and be capable of doing so without seeing any target examples during pre-training. Relative to other modalities, in time series, we expect that time-based and frequency-based representations of the same example are located close together in the time-frequency space. To this end, we posit that time-frequency consistency (TF-C) -- embedding a time-based neighborhood of an example close to its frequency-based neighborhood -- is desirable for pre-training. Motivated by TF-C, we define a decomposable pre-training model, where the self-supervised signal is provided by the distance between time and frequency components, each individually trained by contrastive estimation. We evaluate the new method on eight datasets, including electrodiagnostic testing, human activity recognition, mechanical fault detection, and physical status monitoring. Experiments against eight state-of-the-art methods show that TF-C outperforms baselines by 15.4% (F1 score) on average in one-to-one settings (e.g., fine-tuning an EEG-pretrained model on EMG data) and by 8.4% (precision) in challenging one-to-many settings (e.g., fine-tuning an EEG-pretrained model for either hand-gesture recognition or mechanical fault prediction), reflecting the breadth of scenarios that arise in real-world applications. Code and datasets: https://github.com/mims-harvard/TFC-pretraining.

Rapid and versatile object manipulation in air is an open challenge. An energy-efficient and adaptive soft gripper combined with an agile aerial vehicle could revolutionize aerial robotic manipulation in areas such as warehousing. This paper presents a bio-inspired gripper powered by hydraulically amplified electrostatic actuators mounted to a quadcopter that can interact safely and naturally with its environment. Our gripping concept is motivated by an eagle's foot. Our custom multi-actuator concept is inspired by a scorpion tail design (consisting of a base electrode with pouches stacked adjacently) and spider-inspired joints (classic pouch motors with a flexible hinge layer). A hybrid of these two designs realizes a higher force output under moderate deflections of up to 25{\deg} compared to single-hinge concepts. In addition, sandwiching the hinge layer improves the robustness of the gripper. For the first time, we show that soft manipulation in air is possible using electrostatic actuation. This study demonstrates the potential of untethered hydraulically amplified actuators in aerial robotic manipulation. Our proof of concept opens up the use of hydraulic electrostatic actuators in mobile aerial systems.

Zurich - Founded in 2021, medtech aiEndoscopic combines artificial intelligence (AI) with robotic endoscopy. Its first development, intuBot, presents an all-in-one solution for airway management. It aims to make the life-saving process of endotracheal intubation, used in emergency medicine in particular, much easier and thus save lives. It is based on a collaborative project with ETH Zurich, University Hospital Zurich, and the University of Zurich. Additionally, the startup has been selected to form part of the Switzerland Innovation Park Zürich as well as to join the group of Swiss Venture Leaders Medtech 2022.

Despite these severe impacts, buildings have one of the most significant potentials for green transformation. Today, data-driven technology empowered by machine learning models or artificial intelligence (AI) creates smart buildings – where the software automatically integrates with the building's components. HVAC, air quality, temperature, energy use, occupancy, downtime hours, ventilation, and many other factors can be continually monitored with sensors paired with monitoring technology and can make automated decisions to optimize performance.

Unsupervised mixture learning (UML) aims at identifying linearly or nonlinearly mixed latent components in a blind manner. UML is known to be challenging: Even learning linear mixtures requires highly nontrivial analytical tools, e.g., independent component analysis or nonnegative matrix factorization. In this work, the post-nonlinear (PNL) mixture model -- where unknown element-wise nonlinear functions are imposed onto a linear mixture -- is revisited. The PNL model is widely employed in different fields ranging from brain signal classification, speech separation, remote sensing, to causal discovery. To identify and remove the unknown nonlinear functions, existing works often assume different properties on the latent components (e.g., statistical independence or probability-simplex structures). This work shows that under a carefully designed UML criterion, the existence of a nontrivial null space associated with the underlying mixing system suffices to guarantee identification/removal of the unknown nonlinearity. Compared to prior works, our finding largely relaxes the conditions of attaining PNL identifiability, and thus may benefit applications where no strong structural information on the latent components is known. A finite-sample analysis is offered to characterize the performance of the proposed approach under realistic settings. To implement the proposed learning criterion, a block coordinate descent algorithm is proposed. A series of numerical experiments corroborate our theoretical claims.

Modern Lightweight robots are constructed to be collaborative, which often results in a low structural stiffness compared to conventional rigid robots. Therefore, the controller must be able to handle the dynamic oscillatory effect mainly due to the intrinsic joint elasticity. Singular perturbation theory makes it possible to decompose the flexible joint dynamics into fast and slow subsystems. This model separation provides additional features to incorporate future knowledge of the jointlevel dynamical behavior within the controller design using the Model Predictive Control (MPC) technique. In this study, different architectures are considered that combine the method of Singular Perturbation and MPC. For Singular Perturbation, the parameters that influence the validity of using this technique to control a flexible-joint robot are investigated. Furthermore, limits on the input constraints for the future trajectory are considered with MPC. The position control performance and robustness against external forces of each architecture are validated experimentally for a flexible joint robot.

We construct and analyze approximation rates of deep operator networks (ONets) between infinite-dimensional spaces that emulate with an exponential rate of convergence the coefficient-to-solution map of elliptic second-order partial differential equations. In particular, we consider problems set in $d$-dimensional periodic domains, $d=1, 2, \dots$, and with analytic right-hand sides and coefficients. Our analysis covers linear, elliptic second order divergence-form PDEs as, e.g., diffusion-reaction problems, parametric diffusion equations, and elliptic systems such as linear isotropic elastostatics in heterogeneous materials. We leverage the exponential convergence of spectral collocation methods for boundary value problems whose solutions are analytic. In the present periodic and analytic setting, this follows from classical elliptic regularity. Within the ONet branch and trunk construction of [Chen and Chen, 1993] and of [Lu et al., 2021], we show the existence of deep ONets which emulate the coefficient-to-solution map to a desired accuracy in the $H^1$ norm, uniformly over the coefficient set. We prove that the neural networks in the ONet have size $\mathcal{O}(\left|\log(\varepsilon)\right|^\kappa)$, where $\varepsilon>0$ is the approximation accuracy, for some $\kappa>0$ depending on the physical space dimension.