In manufacturing processes, surface inspection is a key requirement for quality assessment and damage localization. Due to this, automated surface anomaly detection has become a promising area of research in various industrial inspection systems. A particular challenge in industries with large-scale components, like aircraft and heavy machinery, is inspecting large parts with very small defect dimensions. Moreover, these parts can be of curved shapes. To address this challenge, we present a 2-stage multi-modal inspection pipeline with visual and tactile sensing. Our approach combines the best of both visual and tactile sensing by identifying and localizing defects using a global view (vision) and using the localized area for tactile scanning for identifying remaining defects. To benchmark our approach, we propose a novel real-world dataset with multiple metallic defect types per image, collected in the production environments on real aerospace manufacturing parts, as well as online robot experiments in two environments. Our approach is able to identify 85% defects using Stage I and identify 100% defects after Stage II.

Equipment failures can cause major disruptions to system operations. Although this is the case for engineered systems in general, it is especially applicable to autonomous systems as an operation or maintenance crew may not be available to remediate the situation during operation. Autonomous vehicles, for instance, may be performing critical missions miles away from the nearest manned support personnel when a failure occurs. In this work we propose and test an approach for automated predictive reconfiguration of an autonomous vehicle with the goal of delaying the occurrence of failures that would otherwise compromise mission accomplishment. The proposed approach is based on the Monte Carlo Tree Search (MCTS) method and assumes the availability of models describing relevant failure mechanisms and the relation between degradation and performance for each failure mode. Our solution introduces novel means for taking into account the uncertainty resulting from estimation of relevant parameters and states , with benefits in terms of reduction of computational cost compared to existing solutions. The proposed approach is successfully tested in a simulation environment.

We consider the problem of estimating states ( e.g., position and velocity) and physical parameters ( e.g., friction, elasticity) from a sequence of observations when provided a dynamic equation that describes the behavior of the system. The dynamic equation can arise from first principles ( e.g., Newton’s laws) and provide useful cues for learning, but its physical parameters are unknown. To address this problem, we propose a model that estimates states and physical parameters of the system using two main components. First, an autoencoder compresses a sequence of observations ( e.g., sensor measurements, pixel images) into a sequence for the state representation that is consistent with physics by including a simulation of the dynamic equation. Second, an estimator is coupled with the autoencoder to predict the values of the physical parameters. We also theoretically and empirically show that using Fourier feature mappings improves the generalization of the estimator in predicting physical parameters compared to raw state sequences when learning from high-frequency data. In our experiments on three visual and one sensor measurement tasks, our model imposes interpretability on latent states and achieves improved generalization performance for long-term prediction of system dynamics over state-of-the-art baselines.

We consider the problem of reinforcement learning when provided with (1) a baseline control policy and (2) a set of constraints that the learner must satisfy. The baseline policy can arise from demonstration data or a teacher agent and may provide useful cues for learning, but it might also be sub-optimal for the task at hand, and is not guaranteed to satisfy the specified constraints, which might encode safety, fairness or other application-specific requirements. In order to safely learn from baseline policies, we propose an iterative policy optimization algorithm that alternates between maximizing expected return on the task, minimizing distance to the baseline policy, and projecting the policy onto the constraint-satisfying set. We analyze our algorithm theoretically and provide a finite-time convergence guarantee. In our experiments on five different control tasks, our algorithm consistently outperforms several state-of-the-art baselines, achieving 10 times fewer constraint violations and 40% higher reward on average.

We consider the problem of learning control policies that optimize a reward function while satisfying constraints due to considerations of safety, fairness, or other costs. We propose a new algorithm, Projection-Based Constrained Policy Optimization (PCPO). This is an iterative method for optimizing policies in a two-step process: the first step performs a local reward improvement update, while the second step reconciles any constraint violation by projecting the policy back onto the constraint set. We theoretically analyze PCPO and provide a lower bound on reward improvement, and an upper bound on constraint violation, for each policy update. We further characterize the convergence of PCPO based on two different metrics: $\normltwo$ norm and Kullback-Leibler divergence. Our empirical results over several control tasks demonstrate that PCPO achieves superior performance, averaging more than 3.5 times less constraint violation and around 15\% higher reward compared to state-of-the-art methods.

Editor’s note: Data-driven design methods are a promising way to design computing systems at various scales. This article presents a survey regarding data-driven techniques to design sustainable computing systems. —Partha Pratim Pande, Washington State University

The controlling method uses a control specification. In said method, at least part of the progression of the controlling process is monitored, and at least one quality criterion characterizing the quality of the control method is determined, i.e. ascertained, in accordance with said progression. The control specification is adjusted in accordance with the quality criterion.

We are pleased to present the special issue on “Smart and Autonomous Systems for Sustainability: Sustainable Computing and Computing for Sustainability.” We are witnessing the rise of the data-driven science paradigm, in which massive amounts of streaming data—much of it collected as a side-effect of ordinary human activity—can be analyzed to make sense and be able to make intelligent decisions for sustainability over multiple time scales (e.g., short-term versus long-term planning). Intuitively, sustainability refers to the ability to maintain a certain performance or efficiency over time. This concept in the context of computing systems can be seen as sustainable computing—computing systems that maintain a certain performance reliably by consuming low power for a very long period of time. To enable sustainable computing, we need adaptive approaches for managing the computing resources and methods for improving reliability and security of computing systems.

A method of operating an intelligent programmable logic controller (PLC) as part of a production process within an automation system includes the intelligent PLC receiving automation system data and a semantic context model comprising a plurality of ontologies providing formal specifications of conceptual entities associated with the automation system. The intelligent PLC creates one or more semantic annotations for the automation system data using the semantic context model. These semantic annotations are stored along with the automation system data in a non-volatile storage medium included in the intelligent PLC.

The effects of realistic wireless communication are important for the modeling and evaluation of emerging connected and automated vehicle applications. This paper discusses the development of a novel closedloop connected vehicle analysis system (CONVAS) interlinking Vissim microscopic traffic simulation and ns-3 wireless communication simulation. In CONVAS, ns-3 uses Vissim's vehicle positions for wireless modeling and in turn Vissim uses the packet reception from ns-3 to modify vehicle behaviors. The authors propose an application for intelligent avoidance of the dilemma zone to demonstrate the use of the platform for modeling connected and automated vehicles and to illustrate the effects of communications on the application performance. A simulation experiment was conducted with a test bed for an isolated highspeed, fixed-time signalized intersection with different communication model configurations. The evaluation results showed that the effectiveness of the application as measured by successful clearing of vehicles from the dilemma zone depended greatly on performance of wireless communication.

Railway vehicles are generally maintained preventively within certain time periods. Condition based predictive maintenance strategies have a great economic potential so that modern trains are equipped with many sensors in order to perform diagnostics and prognostics of components.Methods for fault detection need appropriate feature subsets in order to achieve small in-sample and out-sample errors. In our case the typical feature selection approach using pure data-driven methods is difficult, as the number of possible feature sets is very large. On the other hand there exists rich domain knowledge and detailed physical models of the mechanical system. The aim is to combine this knowledge with the often used mathematical methods for feature selection for improving classification of cases when a faulty damper is present. Based on the dynamic equations of motion, this paper presents heuristic feature selection via the analysis of transfer functions. We describe several wellknown methods of automated feature selection and a workflow which combines domain knowledge with automated methods. Results show that it is difficult to definefeatures based only on domain-knowledge, but in combination with data-driven techniques good classification performance can be achieved.

We extend the maximum likelihood method for wideband direction of arrival (DOA) estimation to the case of an unknown number of moving sources. The extension is nontrivial because closed-form expressions for the conditional signal covariance matrices are no longer available. We propose a reversible jump particle filter (RJPF) based estimation of the source angles, which has been successfully used in narrowband DOA estimation of moving sources. We discuss added difficulties in DOA estimation compared to frequency retrieval problems. These difficulties are addressed by appropriate modifications of the underlying stochastic model. Finally, we show how an estimator of the number of sources and their positions can be constructed from a discrete representation of their posterior probabilities as provided by the particle filter.

If accurate short term prediction of electricity consumption is available, the Smart Grid infrastructure can rapidly and reliably react to changing conditions. The economic importance of accurate predictions justifies research for more complex forecasting algorithms. This paper proposes road traffic data as a new input dimension that can help improve very short term load forecasting. We explore the dependencies between power demand and road traffic data and evaluate the predictive power of the added dimension compared with other common features, such as historical load and temperature profiles.