Driven by the increasing demand for accurate and efficient representation of 3D data in various domains, point cloud sampling has emerged as a pivotal research topic in 3D computer vision. Recently, learning-to-sample methods have garnered growing interest from the community, particularly for their ability to be jointly trained with downstream tasks. However, previous learning-based sampling methods either lead to unrecognizable sampling patterns by generating a new point cloud or biased sampled results by focusing excessively on sharp edge details. Moreover, they all overlook the natural variations in point distribution across different shapes, applying a similar sampling strategy to all point clouds. In this paper, we propose a Sparse Attention Map and Bin-based Learning method (termed SAMBLE) to learn shape-specific sampling strategies for point cloud shapes. SAMBLE effectively achieves an improved balance between sampling edge points for local details and preserving uniformity in the global shape, resulting in superior performance across multiple common point cloud downstream tasks, even in scenarios with few-point sampling.

Contrastive learning is an essential method in self-supervised learning. It primarily employs a multi-branch strategy to compare latent representations obtained from different branches and train the encoder. In the case of multi-modal input, diverse modalities of the same object are fed into distinct branches. When using single-modal data, the same input undergoes various augmentations before being fed into different branches. However, all existing contrastive learning frameworks have so far only performed contrastive operations on the learned features at the final loss end, with no information exchange between different branches prior to this stage. In this paper, for point cloud unsupervised learning without the use of extra training data, we propose a Contrastive Cross-branch Attention-based framework for Point cloud data (termed PoCCA), to learn rich $3 D$ point cloud representations. By introducing sub-branches, PoCCA allows information exchange between different branches before the loss end. Experimental results demonstrate that in the case of using no extra training data, the representations learned with our self-supervised model achieve state-of-the-art performances when used for downstream tasks on point clouds.

The OPC UA standard combines industrial communication with information modeling. In order to remove the need for continuous polling, OPC UA clients can create subscriptions to be notified about data changes and events. OPC UA Event Filters provide powerful constructs for the server-side selection of relevant events. This feature is however underused in practice, also because of the complexity of Event Filters. Event Filters are expressed in a “byte-code” encoding. Assembling this encoding by hand is difficult and error-prone. This paper discusses the semantics of OPC UA Event Filters and proposes a query language to facilitate their definition. A formal grammar for the query language is provided in the BNF format. An implementation of a corresponding parser is available as part of the open62541 library.

In cyber-physical production systems, the capabilities and states of field resources are continuously available. For adaptive and flexible production control, this information must be evaluated during process execution, which is often not done due to a static job shop schedule. With the presented Execution Engine, we demonstrate a system architecture for order-oriented, fully automated execution of production orders. For the representation of field resources and the interface to the production process control, the OPC UA standard is utilized. Based on an enriched description of the order production steps, an execution considering the states of individual field components is realized. The increased communication overhead is addressed by an OPC UA Aggregation Device Registry, which enables efficient pre-selection of field resources during runtime allocation.

Manufacturing process optimization is an open question, where Bayesian decision theoretic methods have shown considerable promise. One such is Bayesian optimization, with Gaussian Process (GP) surrogate model. This paper explores Gaussian Processes networks to jointly use parameter and observed state to predict the output(s) of a manufacturing process. The Gaussian process network that represents the paths from parameters to state-space to tasks, provides a methodology to ‘look inside’ the black-box of complex manufacturing processes. We present a comparative analysis of this method against the multi-task Gaussian processes and single-task counterparts, highlighting the benefits and drawbacks of each in modelling the behavior of such processes. We show the benefits of the proposed approach using numerical experiments. We show that we are able to improve the output prediction by additional sensor observations from inside the process at training time without needing those sensor observations for predicting product quality given the process parameters.

We propose a combined generative and contrastive neural architecture for learning latent representations of 3D volumetric shapes. The architecture uses two encoder branches for voxel grids and multi-view images from the same underlying shape. The main idea is to combine a contrastive loss between the resulting latent representations with an additional reconstruction loss. That helps to avoid collapsing the latent representations as a trivial solution for minimizing the contrastive loss. A novel dynamic switching approach is used to cross-train two encoders with a shared decoder. The switching approach also enables the stop gradient operation on a random branch. Further classification experiments show that the latent representations learned with our self-supervised method integrate more useful information from the additional input data implicitly, thus leading to better reconstruction and classification performance.

Point cloud sampling is a less explored research topic for this data representation. The most commonly used sampling methods are still classical random sampling and farthest point sampling. With the development of neural networks, various methods have been proposed to sample point clouds in a task-based learning manner. However, these methods are mostly generative-based, rather than selecting points directly using mathematical statistics. Inspired by the Canny edge detection algorithm for images and with the help of the attention mechanism, this paper proposes a non-generative Attention-based Point cloud Edge Sampling method (APES), which captures salient points in the point cloud outline. Both qualitative and quantitative experimental results show the superior performance of our sampling method on common benchmark tasks.

The safety validation of AI and ML-based systems is challenging, as (i) analytical validation needs to include the interaction with a complex and stochastic physical environment and (ii) empirical validation needs to observe very long time-horizons to get enough “statistical signal” for the typically very low safety-related incident rate. This paper proposes an approach that amplifies the empirical evidence by introducing a handicap that reduces the system performance—making safety-related failures empirically more visible in a controlled environment—and gradually removing the handicap so that the convergence to the final incident rate can be estimated. Two numerical case studies are used to support and exemplify the approach.

Anomalies in production processes can cause expensive standstills, damages to the production equipment, waste of materials and flaws in the final product. In production, finding anomalies is usually accomplished by machine learning methods. But to avert anomalies and to automatically recover, actually the detection of the root causes is required. We developed an approach that detects anomalies and then deduces root causes by combining an anomaly detector with a novel Root Cause Analysis (RCA) method based on a causal graph. This specific combination of methods allows causally justified, explainable and counterfactual RCA. The developed algorithm was applied to a simulated gripping process using robotic arms. It found the two root causes of the detected anomalies in the simulated scenarios.

On robotics computer vision tasks, generating and annotating large amounts of data from real-world for the use of deep learning-based approaches is often difficult or even impossible. A common strategy for solving this problem is to apply simulation-to-reality (sim2real) approaches with the help of simulated scenes. While the majority of current robotics vision sim2real work focuses on image data, we present an industrial application case that uses sim2real transfer learning for point cloud data. We provide insights on how to generate and process synthetic point cloud data in order to achieve better performance when the learned model is transferred to real-world data. The issue of imbalanced learning is investigated using multiple strategies. A novel patch-based attention network is proposed additionally to tackle this problem.

Despite its great success, machine learning can have its limits when dealing with insufficient training data. A potential solution is the additional integration of prior knowledge into the training process which leads to the notion of informed machine learning. In this paper, we present a structured overview of various approaches in this field. We provide a definition and propose a concept for informed machine learning which illustrates its building blocks and distinguishes it from conventional machine learning. We introduce a taxonomy that serves as a classification framework for informed machine learning approaches. It considers the source of knowledge, its representation, and its integration into the machine learning pipeline. Based on this taxonomy, we survey related research and describe how different knowledge representations such as algebraic equations, logic rules, or simulation results can be used in learning systems. This evaluation of numerous papers on the basis of our taxonomy uncovers key methods in the field of informed machine learning.

OPC UA is a client-server communication protocol for industrial use cases without hard realtime requirements. The new PubSub extension of OPC UA adds the possibility of many-to-many communication based on the Publish / Subscribe paradigm. In conjunction with the upcoming Time-Sensitive Networking (TSN) extensions of Ethernet, OPC UA Pub Sub aims to also cover time-deterministic connectivity. This poses requirements to OPC UA implementations that have traditionally not been regarded. We propose an approach to combine non-realtime OPCUA servers with realtime OPC UA Pub Sub where both can access a shared information model without the loss of realtime guarantees for the publisher. As a result, the publisher can be run inside a (hardware-triggered) interrupt to ensure short delays and small jitter. An open source implementation of OPC UA Pub Sub is provided based on the open62541 SDK. This is also the basis for measurements used to evaluate the potential of the technology.

Automated manufacturing systems are becoming increasingly flexible in order to support a growing number of different products and product variations, as well as shortening lot sizes and product life cycles. Many manufacturing resource are already multi-purpose. But they are integrated in an automation infrastructure that may require significant effort to adapt. In this contribution, we present a system architecture for the deployment of software functionality to manufacturing resources safely at runtime. The architecture integrates software model verification to ensure the integrity of new functionality, software-defined networking (SDN) to establish new communication pathways for collaboration, and the close interaction between a flexible runtime execution environment and its safety-critical counterpart.

OPC UA and DDS are communication protocols for the Industrial Internet. However, they make use of contrasting communication patterns and represent different architectural styles. We discuss these differences and their impact in an Industrial Internet and Internet of Things context. Further, we show up the possibilities for hybrid implementations leveraging the features of both OPC UA and DDS. For this, we provide a) a mapping of the OPC UA data types into DDS and b) a set of DDS quality of service policies that match the guarantees made by the standard OPC UA binary protocol.

Representational state transfer (REST) is a wide-spread architecture style for decentralized applications. REST proposes the use of a fixed set of service interfaces to transfer heterogeneous resource representations instead of defining custom interfaces for individual applications. This paper explores the advantages of RESTful architectures, i.e., service-oriented software architectures comprised RESTful services, in industrial settings. These include communication advantages such as reduced communication overhead and the possibility to introduce caching layers, and system design advantages including stable service interfaces across applications and the use of resource-oriented information models in cyber-physical systems. Additionally, a RESTful extension to the open platform communications (OPC) unified architecture (OPC UA) binary protocol is proposed in order to leverage these advantages. It requires only minimal modifications to the existing OPC UA stacks and is fully backward compatible with the standard protocol. Performance benchmarks on industrial hardware show a throughput increase up to a factor of eight for short-lived interactions. This reduction of overhead is especially relevant for the use of OPC UA in the emerging Industrial Internet of Things.

As a standardized communication protocol, OPC UA is the main focal point with regard to information exchange in the ongoing initiative Industrie 4.0. But there are also considerations to use it within the Internet of Things. The fact that currently no open reference implementation can be used in research for free represents a major problem in this context. The authors have the opinion that open source software can stabilize the ongoing theoretical work. Recent efforts to develop an open implementation for OPC UA were not able to meet the requirements of practical and industrial automation technology. This issue is addressed by the open62541 project which is presented in this article including an overview of its application fields and main research issues.

RESTful interfaces are a wide-spread architecture style for webservice implementations and are built upon the resource-oriented approach to decentralized architectures (ROA). REST postulates a set of requirements that are not covered by the OPC Unified Architecture (OPC UA) communication protocol per se. We propose a set of simple extensions to the OPC UA binary protocol that enable RESTful communication. The evaluation shows an order of magnitude improvement in the use of communication resources for sporadic service requests. Additionally, RESTful OPC UA allows applications to profit from the advantages of the resource-oriented architecture style, such as caching and loose application coupling.

Shortening product lifecycles and small lot sizes require manufacturing systems to adapt increasingly fast. Many existing machine tools, handling and logistics systems are already generic and not bound to a specific product a-priori. Yet this flexibility and reconfigurability on the asset level is lost in automated systems that are limited to executing a small set of predefined actions in a fixed sequence. The SkillPro1project aims to develop a holistic service-oriented framework for modelling and orchestration of modern adaptable manufacturing systems. The core concept is a unified abstraction for manufacturing tasks: skills provided by the available assets and the requirements of the different production steps. The skill-based system model enables the transition from generic high-level descriptions to low-level formats that can be directly executed. Self-describing assets can be added, changed and removed at runtime, taking into account technical and economic conditions to best achieve the manufacturing goals.

This paper considers the efficient operation of shared mobility systems via the combination of intelligent routing decisions for staff-based vehicle redistribution and real-time price incentives for customers. The approach is applied to London's Barclays Cycle Hire scheme, which the authors have simulated based on historical data. Using model-based predictive control principles, dynamically varying rewards are computed and offered to customers carrying out journeys, based on the current and predicted state of the system. The aim is to encourage them to park bicycles at nearby underused stations, thereby reducing the expected cost of redistributing them using dedicated staff. In parallel, routing directions for redistribution staff are periodically recomputed using a model-based heuristic. It is shown that it is possible to trade off reward payouts to customers against the cost of hiring staff to redistribute bicycles, in order to minimize operating costs for a given desired service level.

To model increasingly adaptive production systems, skills are used to describe generic capabilities of the system components. In this paper, the authors extend the well-known division of production entities into product, process, and resource (PPR) with a skill definition. There are two main advantages for this approach: First, using PPR for the skill definition allows easy integration into existing models and tools. Second, there is a natural tendency to define very generic skills to capture all possible use cases. But at some point, skills have to be translated into precise instructions for execution. The model makes this dichotomy explicit and provides a common taxonomy for stakeholders concerned with skills on different abstraction levels.