The aim of this Experiment is to improve the existing model maintained by the digital twin, by having robotic agents making sense of their surrounding world in order to detect inconsistencies between the real world and the digital representation. It focuses on dynamically allocating production resources to manufacturing tasks to better face uncertainties like machine failures and unavailability of operators.

The aim of the Experiment is to improve the services to end-user customers, optimize the maintenance process, and improve the quality control of the measurements. The values of groundwater level, battery state, and other parameters are logged by their data logger system and sent at pre-programmed intervals to the remote server. There is additional storage of maintenance procedures in addition to measurement data. Due to the large amount of data collected, a manual inspection of the collected data is complex. Thus, inevitably it requires an automatic procedure to perform the quality control tests. The AI model will drive the information extraction from the correlation between the measurement and the maintenance data to perform automatic quality control and failure detection system. The AI system will assist the maintenance staff by reducing the maintenance effort and increasing trust in the measurement platform.

The aim of this Experiment is to demonstrate a new decision support tool to aid the manufacturing system designer and reconfiguration planner during greenfield and brownfield design scenarios. Designing a manufacturing environment (workstation, cell, line) is challenging, multi-faceted, and time-consuming task for a human. Currently, designers find feasible resource solutions by comparing the characteristics of the product to the technical properties of the available resources by browsing through online or paper catalogues to select thousands of components manually. A lot of time and knowledge is needed to browse and find components fulfilling the required manufacturing processes from these various catalogues. Yet another question is to integrate the found resources all together so that the system can be connected and will play along. The proposed AI solution improves situation by finding feasible resource combinations out of a large set of available resources and proposes these solutions for the designer. The designer can then focus on comparing only feasible and working solution alternatives together, and to select and configure the final system solution out of them. The experimented capability matchmaking system is expected to help the system designer or reconfiguration planner to find and utilize resources and resource combinations that are out of his/her previous solution space, thus leading to more innovative system solutions. Using new computer-aided intelligent planning methods and tools, the time and effort put into system design can be reduced drastically.

The aim of the Experiment is to improve the heat demand prediction and valve settings in the substations, based on the weather predictions of the district heating (DH) system of Stadsverwarming Purmerend. The weather conditions (apart from the behaviour of the users / clients of the DH) are one of the main variables used to plan demand of heat temperature and temperatures across the distribution network. Such an approach requires a very accurate estimation of needed heat for safe securing their operation in advance. In other words, they need accurate prediction for planning not only energy supply, but all other activities connected with this one as its purchase and preparing technology as well. The DH at Purmerend has a historical data from 10 years of operations. Moreover, the forecast of weather conditions is a non-linear problem by nature. Additional to that, the weather conditions are not limited to only temperature, but to humidity, precipitation and wind speed. Thus, due the multivariate characteristics of data, the seasonality, the need to forecast the weather conditions into different time basis, from minutes to days ahead, taking also into account the lag time in the heat distribution, the need of AI is imperative in this kind of problem, to increase the efficiency of Purmerend DH system.

Recently, new technologies are entering the industrial market, such as Augmented Reality (AR). In these systems, the employee is guided step-by-step by means of projected instructions. The current systems are static: all operators receive similar instructions. However, there is a great need to make such systems adaptive and self-learning using AI: Dynamic customization of operator guidance based on skill level, performance, errors, instantaneous operator capacity and operator knowledge and feedback to improve effectivity of instructions. To address these issues, it is necessary to get insight in the right parameters. The aim of the Experiment is to work towards objective data acquisition from dynamic effects in the manufacturing process, for instance learning curve, fatigue, and operator feedback.

The aim of the Experiment is to provide end-users Small and medium-sized enterprises (SMEs) an instrument to implement AI and AI-related technologies into their production environment. Benefits for end-users can be achieved in terms of time and costs reduction, final product quality and general better production plant management, with the possibility to allow interaction between the production line management and other services (e.g. predictive maintenance, digital twins...). A better production line management will also allow for productivity enhancement and reduction of dead costs, such as warehouse management and waste management. Thanks to AI enhanced predictions, activities inside the production environment will be fine-tuned, allowing for a better efficiency and cost reduction. More, the AI-enhanced management tool will provide support in terms of energy efficiency and machine allocation. Training algorithms on different datasets from different experiences will allow the acquisition of methodologies from other production sectors, fostering cross-contamination.

The aim of this Experiment and the proposed solutions is to reduce costs directly linked to the machining tooling. These savings are granted through the extension of the tooling usage and even its lifetime through the usage of AI techniques. At the same time, once the tooling usage is maximized, the saving costs will come from the reduction on number of operators required to manage the lines, that is every operator will be able to manage more machines because the period between tooling changes will be longer. Additionally, the availability of lines will increase and so the production.

The aim of the Experiment is to implement a system able to support the operator in the identification of defects and anomalies / faulty pieces. The resolution of technical problems and the automatic update of failure dataset can also have a considerable impact on the overall production planning. In the age of Industry 4.0, machine and deep learning has attracted increasing interest for various research applications. In recent years, such AI models have been extensively implemented in fault detection and diagnosis systems. The machine architecture’s automated feature learning process offers great potential to solve problems with traditional fault detection and diagnosis systems. The latter rely on manual feature selection, which requires prior knowledge of the data and is time intensive. However, the high performance of machine/ deep learning comes with challenges and costs.

The aim of this Experiment is to demonstrate the feasibility of introducing Artificial Intelligence techniques in the Predictive Maintenance of stamping presses through the collaboration of: A Technology Center that develops and implements AI algorithms. A Predictive Maintenance service provider that will implement AI in its monitoring and diagnostic tools. An automobile manufacturing company (end user) that makes its stamping facilities available for the Experiment. other production sectors, fostering cross-contamination. The general objective of the Experiment is to contribute to the improvement of the availability and performance of the machines in the stamping sector. This will significantly improve the OEE (Overall Equipment Effectiveness) in the sector, currently between 70-80%. These values compromise the viability of the businesses, making the competitiveness of many companies in the sector low.

The Experiment goal is the implementation of an AI tool that defines and optimized the machine parameters (rotation and linear speed) for a future product considering historic data and product dimensions and focusing on energy consumption reduction. The expected solution will consist of an optimization tool on top of the previous predicting model. This optimization tool will provide a set of solutions to minimize the energy consumption of the modelled machine. This solution will prescript the controlling parameters of the machine. In principle, this tool is designed to run offline and in an open loop, as a decision support for the operators. This information will be provided on a human machine interface connected to a web server. The operator nowadays configures the machine according to their experience (AS-IS). The optimization tool will be used by the operator to receive information about the optimal setpoint that will minimize the energy consumption per reference.