Machine Learning by ZENNER AI

The Challenge

Anomalies has to be detected in large amounts of IoT data (> 300.000 devices) considering the “device history” represented by messages of the past six months. The analytical application has to learn from the device history and detect also unknown anomalies. The extraction of the raw-events sent by the devices and the data-preparation must be implemented in a highly scalable way in order to handle the future increasing number of devices. For an informative reporting, the model assigns every device to a cluster based on the current and historical device status represented by the events that were sent by the device. The clusters for detailed reporting are grouped into categories that enables the focus only on devices with a specific status.

The Solution

Implementation of machine-learning algorithms (unsupervised learning with K-Means-Clustering) on the ZENNER AI Analytic Platform to cluster mass data for detection of different kinds of anomalies in the sensor data. Usage of Big Data technology to scale with the increasing number of devices and event data. Visualization of the results on different levels of details including a “traffic light system”

• Clustering of the devices into detailed clusters representing the device history
• Detailed analysis by enrichment of the clustered devices with information about the location and firmware-version
• Flexible analysis by using filtering and drill-down even to device level
• Notification in case that the number of devices in spesific clusters significantly increases

The Challenge

Energy providers and industrial enterprises face the challenge of accurately forecasting energy consumption while identifying potential anomalies early on. Often, only limited historical data is available, and there are no established patterns of unusual behavior to rely on. Additionally, generating forecasts for the upcoming day is critical for effective planning and resource allocation. Errors in these predictions can lead to unnecessary costs, inefficient resource usage, and unexpected system failures.

The Solution

By leveraging Machine Learning and Artificial Intelligence, an intelligent solution is developed that analyzes historical data and employs advanced modeling techniques. This solution enables:

• Accurate forecasts of energy consumption for the next day, based on existing data and current conditions.
• Anomaly detection in forecasted data to identify potential deviations from normal patterns.
• Near real-time monitoring, allowing for quick integration of current data and continuous adjustments.
• Visualization of results on a user-friendly platform that supports both technical and non-technical users.

The Challenge

Maintaining critical infrastructure across industries such as e-mobility, desalination plants, and solar energy involves a variety of challenges.

• E-Mobility: Charging stations often face issues such as broken connectors, cooling system failures, connection problems, and faulty card readers. Additionally, data availability for predictive analytics can be limited.
• Desalination Plants: Asset aging, resource inefficiencies, unplanned downtimes, and difficulties in optimizing existing equipment significantly impact performance. Remote monitoring and effective lifecycle management are often lacking.
• Solar Panels: Harsh environmental conditions, such as extreme heat and sandstorms, challenge system durability and sensor reliability. Maintenance scheduling and integration with existing systems further complicate operations.

These challenges lead to higher maintenance costs, reduced operational efficiency, and unscheduled downtime, affecting both service reliability and customer satisfaction.

The Solution

Predictive maintenance powered by Machine Learning (ML) and Artificial Intelligence (AI) offers tailored solutions for these industries:

• E-Mobility: Predictive analytics detect early signs of component failures, optimize cooling system maintenance schedules, and monitor card reader performance in real-time to ensure uninterrupted charging operations.
• Desalination Plants: AI models analyze asset performance to optimize resource usage, predict equipment depreciation, and enable remote monitoring of critical systems, reducing downtime and improving overall efficiency.
• Solar Panels: Advanced algorithms improve data accuracy, forecast sensor failures, and schedule maintenance proactively to mitigate the effects of extreme weather. Seamless integration with existing systems ensures smooth operation.

The Challenge

The increasing demand for e-mobility requires a solid data basis and advanced analyses to optimize charging infrastructures and their tariff structures. Future expansion and investment in the charging infrastructure has to be supported by deep analysis of the different charge points and charging behaviours. Data from different backend providers has to be automatically extracted and integrated into a unified data model for analysis. This enables comparative analysis for the entire charging infrastructure and the application of machine learning and predictive models. The use of advanced analytics, machine learning and artificial intelligence enables insights into charging behaviour and charging infrastructure utilisation. Forecast models predicts the future energy demand on different forecast horizons. However, the prediction models needs to be able to deal with time-series of variable length and missing data.

The Solution

The ZENNER AI Analytics Platform offers a wide range of possibilities for applying advanced analytics and machine learning to e-mobility and charging infrastructure data.

• Automatic extraction and consolidation of data from different backend systems into a harmonized reporting
• Detailed analysis of charging behavior by distinguishing charging and parking times as well as charging duration
• Flexible analysis of energy consumption at different aggregation levels and time horizons
• Classification of charging stations based on user behavior
• Prediction of future energy consumption based on an additive forecasting model

The Challenge

The increasing integration of renewable energy sources, such as wind and solar, has introduced new complexities into energy load balancing. These sources are intermittent by nature, making it challenging to maintain grid stability and ensure efficient energy distribution. Additionally, managing demand peaks and avoiding network congestion are critical to preventing outages and maintaining service reliability. Traditional methods often lack the advanced forecasting capabilities needed to address these dynamic challenges effectively.

The Solution

By deploying advanced AI and Machine Learning (ML) technologies, energy load balancing can be optimized to meet modern grid demands. This solution offers:

• Intermittent Energy Management: AI-driven forecasting models predict fluctuations in renewable energy generation and adjust load distribution in real-time to stabilize the grid.
• Demand Peak Management: Algorithms analyze consumption patterns and redistribute loads during peak demand periods to avoid strain on the network.
• Congestion Mitigation: Predictive analytics identify potential network bottlenecks, enabling proactive adjustments to maintain smooth energy flow.
• Decision Support Systems: Real-time insights provide operators with actionable recommendations to improve decision-making and adapt to dynamic conditions.