ABB (Germany)

The EDGE Breaker represents a significant evolution in substation architecture by integrating Merging Unit (MU) functionality and sensors directly within circuit breakers. This approach streamlines the engineering process, reduces wiring complexity, and enhances maintenance efficiency. By supporting centralised and virtualised protection and control systems, the EDGE Breaker simplifies installation and commissioning, eliminates/reduces need for low-voltage compartments, and facilitates easy replacement. The implementation of a bidirectional fast common data process bus, adhering to IEC61850 or IEC61869 standards, is crucial for this approach. This approach enables real-time data sharing and rapid response to faults, thereby enhancing the overall resilience of the substation. Additionally, the optimised Breaker unit can now support centralised protection architecture with merging unit onboard which can further have back up protection enabled. This approach promises cost savings, simplified operations, and enhanced reliability for future substation designs.

Abstract Industrial demand-side flexibility is essential for power systems with high shares of variable renewables, yet static grid fees and full load hours incentives can suppress flexible electrification by penalizing peaks and rewarding uniform consumption. This paper introduces an open, extensible, regulation-aware optimization platform that integrates market and emission data with a mixed-integer scheduling model of a bi-valent industrial dryer capable of using electricity and gas. The platform is demonstrated through a representative use case from the energy intensive paper manufacturing sector operating under the current reduced static grid fee pursuant. Because the reduction is contingent on meeting a minimum full load hours threshold, the scope for exploiting the bivalent flexibility potential remains limited. As an alternative, a proposed dynamic grid fee regime featuring time-varying low- and high grid fee windows is evaluated. Any reduction of power consumption from the grid during the high grid fee window does not count against meeting the full load hours threshold. With dynamic fees, the optimization concentrates electric operation within low grid fee windows, increasing electrified heat, reducing emissions, and maintaining competitive costs. Notably, during summer months, these windows align well with periods of lower grid carbon intensity, reinforcing cost–emission co-benefits. However, this alignment deteriorates in winter due to higher variability in renewable generation patterns and constrained window timing, limiting incentive effectiveness. The findings demonstrate that tariff parameterization materially shapes realizable industrial flexibility and that regulation-aware optimization can translate latent technical potential into sustained, temporally targeted electrification aligned with system conditions. The platform enables reproducible, policy-relevant scenario analysis and can aid plant operators for scheduling flexible assets and tariff designers in testing tariff parameters against realistic operational responses.

Large language models (LLMs) have gained significant attention for their success in natural language generation and their widespread adoption across various business sectors. In process automation (PA) engineering, LLMs can be utilized to assist with tasks such as PLC programming, software development, and data processing. Despite their advantages, LLMs face challenges like biased responses, non-deterministic outputs and model hallucinations. Therefore, it is essential to validate the responses generated by LLMs before relying on them. This work presents a novel workflow to automatically validate unstructured outputs generated by LLMs used in question answering (Q&A) applications based on the user-specified documents in the PA domain. The Domain of Validity (DoV) of these LLMs is first estimated based on the user-specified documents. The proposed workflow uses the estimated DoV to evaluate the relevance of the outputs generated by the LLM with respect to the user documents. A regular expression approach is included in the workflow to check whether the mentioned tag names in the LLM output, if any, exist in the user-specified documents. Furthermore, an LLM-as-a-judge approach is used then to cross-check the LLM generated outputs with the ground truth information. The proposed workflow enables users to automatically detect invalid LLM outputs. This allows the user to focus on cases that truly require human judgment and act accordingly by re-prompting the LLM, adding documents, or switching models. A fictitious PA example is employed to demonstrate the versatility and benefits of the developed workflow.

Supervised machine learning (ML) has achieved significant outcomes in the industrial domain, dependent on the availability and accuracy of ground truth labels. However, the usual assumption of a ground truth to exist in the data often represents only an idealization of real-world conditions, as data labeling can be subjective and prone to errors, leading to so-called label noise. Such noise can significantly degrade model performance. Although extensive research has identified methods to improve model robustness against label noise, there is a notable absence of generic, reusable frameworks that allow industrial practitioners to systematically assess model robustness. To address this gap, we extend our previous work and propose a model-agnostic framework designed specifically for evaluating robustness against label noise. Our framework incorporates two distinct label noise perturbation mechanisms: an instance-independent symmetric perturber and an instance-dependent one. We demonstrate the utility of our extended framework through empirical evaluations on two industrial datasets using six relevant time series classification methods from the literature. The results highlight the significant vulnerability of supervised ML models to both noise types and underscore the value of our framework in uncovering these robustness limitations.

Alarm floods remain a challenge in industrial operations, potentially overwhelming human operators with excessive alarm notifications during abnormal situations. To address this, alarm flood classification (AFC) methods utilize historical data to classify recurring alarm patterns automatically. However, the practical utility of these methods may be limited by potential degradation in alarm data quality, resulting from sensor faults, communication errors, or detection delays, which can substantially compromise their classification accuracy and reliability. This paper proposes a novel methodology to systematically analyze the robustness of AFC methods against realistic alarm data quality issues. We introduce four distinct perturbations: missing alarms, false alarms, delayed alarm flood detection, and alarm reordering, to replicate real-world alarm data degradation. We evaluate our methodology using a novel alarm dataset derived from the Tennessee-Eastman process, while examining the robustness of six relevant AFC methods from the literature. The results demonstrate significant variations in robustness across different AFC methods and perturbation types, providing insights into their practical reliability under various realistic scenarios.

Supervised machine learning has presented impressive results in the industrial domain - when it can depend on reliable ground truth labels. However, the concept of golden labels often represents only an idealization of reality. Instead, data labeling can be highly subjective and hence susceptible to corruption, such as unintentional errors, resulting in socalled label noise. Since label noise can significantly impair the performance of supervised machine learning models, it can be helpful to assess their robustness to label noise when evaluating existing limitations. Despite the large number of publications offering best practices and techniques for making models more robust to label noise, literature lacks generic tools that would allow data scientists and industrial practitioners to test and compare which technique works best for their given model. To address this gap, we propose a model-agnostic, reusable framework for measuring the robustness of supervised machine learning models to label noise.

Abstract For us to hear, the cochlea encodes sounds into neural signals at synapses of inner hair cells (IHCs) and the auditory nerve with remarkable fidelity. To achieve the high rates of temporally precise synaptic transmission over long periods of time, IHCs employ sophisticated ribbon-type active zones (AZ). In order for us to understand synaptic sound encoding, we need to decipher the underpinning molecular topography of these synapse which had remained challenging due to technological limitations. Here we applied 3-dimensional minimal flux optical nanoscopy to mouse IHC synapses to chart the position of key pre- and postsynaptic proteins with single digit nanometre resolution of imaging. We demonstrate that nanoclusters of channels and interacting proteins govern the topography of AZs and postsynaptic densities (PSDs). We count synaptic proteins, their nanoclusters and determine their spatial organization feeding into computational modelling of AZ function. In conclusion, this study reveals a nanocluster-based molecular AZ and PSD topography, likely serving as functional modules in synaptic sound encoding.

AutomationML (AML), a standard for the interoperable exchange of engineering data in industrial automation, holds promise for facilitating seamless integration and data exchange across diverse industrial systems. By serializing AML to JSON, the format aims to enable a wide array of functionalities associated with JSON. Yet, such adoptions need to be carefully considered in order to have the support of the modeling community. This paper presents the efforts of the AML-JSON working group to define this serialization. It discusses the methodology employed, the technical aspects of the serialization process, and the implications for industrial applications. Furthermore, it highlights the advantages and challenges of adopting a JSON serialization within the context of AML, offering insights for researchers, practitioners, and industry stakeholders seeking to leverage this technology for enhanced data handling and interoperability in industrial automation.

The problem of deriving a model to predict the clinker quality in a cement production plant is considered. The process has a highly complex and nonlinear dynamic behavior, making physics-based first-principle modelling ineffective. A data-driven approach is thus proposed, to obtain an input-output model able to represent the overall system dynamics and to estimate the quality key performance indicators (KPIs). The approach combines a dynamic linear model, to estimate the evolution of the main process variables, with a static nonlinear regression model, to infer the clinker quality KPIs. The parameters of the dynamic model are estimated recursively on-line in a Moving Horizon Estimation fashion, to adapt to time-varying conditions such as the (unmeasured and uncertain) fuel mix. Real-world data collected on a European cement plant are used both to develop the approach and to test its effectiveness.

The escalating global emissions attributable to commercial buildings call for the integration of digital technologies to enhance energy efficiency and occupant comfort. Cyber-physical control systems (CPCS) can collect data from various sensors and actuators in the building to provide valuable insights to building owners. However, the unstructured metadata text fields in these systems pose a challenge in leveraging artificial intelligence and machine learning solutions for building management. This paper proposes a solution using pre-trained large language models for building metadata normalization that addresses inconsistencies in the point text metadata across buildings. The evaluation of this solution is performed using two publicly-available CPCS datasets, revealing its ability to structure the unstructured natural language metadata without fine tuning or training from scratch.