Raffaello Lepratti

Advanced human–machine system for intelligent manufacturing

The use of ontologies has gained more and more interest above all for the knowledge management, e.g. the exchange of professional “know-how”, as reported in various previous papers. Under the pressure of a turbulent international market situation enterprises stress the importance of innovation in manufacturing areas. For instance, due to the drastic growing automation degree of manufacturing systems an intuitive interaction form is required, which enables the shop-floor personnel an active participation to the production without specific technical background, as well as to capture and retrieve systematically knowledge contents arising from the interaction process.The following contribution takes this topic into consideration and proposes an innovative ontology- based approach called ontological filtering system (OFS) based on methods and procedures to formalize natural language contents in a systematic way. By means of a so-called ontological network (ON) generic term forms used in the human–machine interaction (HMI) via natural language could be led back to a set of pre-defined terms. Thus, the ON consists, on the one hand, of a large number of generic natural language terms and, on the other hand, of a set of so-called key terms. The generic terms are defined, classified in semantic categories and chained together per semantic relations for a specific use in a particular domain of discourse. The key terms are used to build information on machine level and, therefore, have a formal definition. Through additional syntax roles and application-specific semantic constrains a systematic access and processing of natural language instructions is accomplished computationally. The proposed concepts have been set up and tested within an experimental testbed. The obtained results show a high system performance and encourage the research team to invest further efforts, in order to validate the system operational performances towards its industrial use at shop-floor level.

Event-based Reactive Production Order Scheduling for Manufacturing Execution Systems ?

Abstract Today, first inter-organizational tracking & tracing systems facilitate timely identification and handling of disruptions along the supply chain. However, these systems typically operate at SCM and ERP level and therefore lack knowledge and control over production processes. This paper bridges the gap between logistics and production IT by proposing a novel architecture for coupling the manufacturing operations as defined in IEC 62264 with an EPCIS-compliant real-time tracking & tracing system. The system leverages logic-based complex event processing for detecting critical disruptions in the supply chain and initiates rescheduling of production. It is shown that in the presence of unexpected events the rescheduling algorithm minimizes delays and inventory costs while avoiding “nervous schedules” caused by frequent changes. As proof of concept a reference implementation of the architecture is realized within a discrete production line.

Semantic-Guided Feature Selection for Industrial Automation Systems

Modern industrial automation systems incorporate a variety of interconnected sensors and actuators that contribute to the generation of vast amounts of data. Although valuable insights for plant operators and engineers can be gained from such data sets, they often remain undiscovered due to the problem of applying machine learning algorithms in high-dimensional feature spaces. Feature selection is concerned with obtaining subsets of the original data, e.g. by eliminating highly correlated features, in order to speed up processing time and increase model performance with less inclination to overfitting. In terms of high-dimensional data produced by automation systems, lots of dependencies between sensor measurements are already known to domain experts. By providing access to semantic data models for industrial data acquisition systems, we enable the explicit incorporation of such domain knowledge. In contrast to conventional techniques, this semantic feature selection approach can be carried out without looking at the actual data and facilitates an intuitive understanding of the learned models. In this paper we introduce two semantic-guided feature selection approaches for different data scenarios in industrial automation systems. We evaluate both approaches in a manufacturing use case and show competitive or even superior performance compared to conventional techniques.

Event-Driven Order Rescheduling Model for Just-In- Sequence Deliveries to a Mixed-Model Assembly Line

Today’s buyer markets and lean supply chains require build-to-order assembly systems with just-in-sequence (JIS) deliveries. Simultaneously production systems have become prone to supply disturbances (i.e. events) that endanger the synchronized delivery of all JIS components to the assembly line. To uphold production sequence stability, rescheduling is frequently required. Current methods, however, make assumptions that are often insufficiently aligned with real-world problems and focus on production issues while neglecting the implications of today’s tight integration of supply chain with production processes. To this end, this contribution derives a general model of a mixed model assembly line. It then proposes and evaluates an event-driven rescheduling model for JIS deliveries. The results indicate that rework due to missing JIS components can be avoided without compromising performance.

Enhancing Interoperability Through the Ontological Filtering System

European industry is in transition process from a mass production industry towards a knowledge-based customer- and service-oriented one. Consequently, networked, knowledge-driven and agile manufacturing systems emerge as key elements towards the future production scenario. In order to automate production tasks also for small production batch and enable uncomplicated system reconfigurations, a holistic knowledge management architecture, which enables knowledge acquisition and retrieval, is needed. Since several heterogeneous technologies and human actors are still involved in the knowledge transfer process interoperability barriers must be overcome. In this paper an ontology-based approach will be described. It enhances Human-Machine Interoperability simplifying the interaction level between them and makes possible automatic knowledge transfer. First trial tests have been performed by means of a natural language interaction system within the automation technology domain.

Intelligent PC-based user control interface for on-line correction of robot programs

Today the automotive sector is dominated by high variety of types and increasing product diversification. Thus, OEM manufacturers reintegrate key technologies back in the enterprise, due to competitive reason and to stabilize leadership in innovation. As an example, state-of-the-art fuel and diesel engine cylinder heads achieve higher complexity and filigrane structure, therefore careful treatment of specimen as well as net-shape machining processes with high quality output are a major requirement in this sector. A key issue is the surface fretting and finishing of engine parts subsequent following foundry procedures. An advanced factory automation system, based on an industrial robot (6-axis joint coordinate), has been developed in order to obtain the described results. On the system apply various measuring procedures, in order to recognize and remove burr formation (particularly with optical technique). A graphical interactive programming system enables simple and user friendly correction at deburring results for individual workpiece types. A possibility was given to the system user, who can easily remove deburring errors through short operator interaction, mainly input of correction values at controller system. Thereby, each authorized robot cell operator can contribute to guaranty process quality without special technical training.

Knowledge Fusion of Manufacturing Operations Data Using Representation Learning

Due to increasingly required flexibility in manufacturing systems, adaptation of monitoring and control to changing context such as reconfiguration of devices becomes more important. Referring to the usage of structured information on the Web, digital twin models of manufacturing data can be seen as knowledge graphs that constantly need to be aligned with the physical environment. With a growing number of smart devices participating in production processes, handling these alignments manually is no longer feasible. Yet, the growing availability of data coming from operations (e.g. process events) and contextual sources (e.g. equipment configurations) enables machine learning to synchronize data models with physical reality. Common knowledge graph learning approaches, however, are not designed to deal with both, static and time-dependent data.

Graph-based predictions and recommendations in flexible manufacturing systems

Due to the emerging paradigm of mass-customization, manufacturing processes are becoming increasingly complex. Management of this complexity requires system support that goes beyond traditional MES capabilities, such as discovery of patterns throughout massive networks of interdependent processes. As of today, Manufacturing Analytics offer only limited decision support focused on descriptive metrics that cannot account for predictive and prescriptive decision support, such as detection of systematic fault patterns. The application of predictive models in manufacturing environments is non-trivial, because they need to reflect system domain constraints and preserve semantics of manufacturing operations. Recent approaches of so-called Advanced Manufacturing Analytics try to fill this gap by applying standard data mining algorithms with customized data preparation for domain-specific use cases. In order to overcome the problem of high customization efforts, we introduce a graph-based analytics framework derived from a comprehensive requirements analysis. Additionally, we demonstrate applicability of the presented framework on two exemplary manufacturing analytics use cases.

Dynamic cycle times for adaptive manufacturing control in automotive flow shops

The automotive industry has to deal with an increasing complexity of production processes and various kinds of disturbances along the supply chain. This requires a higher level of flexibility through an intelligent production planning from the engineering phase. Todays manufacturing flexibility is mostly provided during the operational phase (so-called runtime) by methodologies for order re-scheduling and re-sequencing. The focus of this paper is a novel concept, which adds the intelligent production planning to these methodologies and uses the synergies of the holistic system. This approach enables flexible automated manufacturing processes by the dynamic use of machine capabilities during run-time. The paper shows in details how the adaption of operating speeds both in manufacturing and material handling processes leads to dynamic cycle times with maximized Key Performance Indicators (KPIs). This concept is based on so-called production variants defined and validated during the engineering phase. First results show stability and good response of the test system.

More reliability in Supply Chain Management

The increasing amount of outsourcing in car manufacturing means that original equipment manufacturers (OEMs) are becoming highly dependent on the supply chain, potentially leading to a loss of control and loss of market transparency. In this context this article from Siemens shows that successful interaction between automation and IT is increasingly important to ensure smooth data management for just-in-time or just-in-sequence processes.