However, the data is only useful if it can be found and understood - insight engines ensure a uniform view of different types of information. Regardless of whether it comes from a cloud, third-party applications or networks.

Elevators, conveyor belts, construction machinery and other systems around the world stand idle for several million hours every year. The reason for this is possible defects, which are not only annoying, but above all very costly. In addition to the loss of revenue caused by business downtime, there are also the usually not inconsiderable repair costs for the defective machine, which could be avoided or at least minimized with optimally planned maintenance.

These costs can be reduced through predictive maintenance. Sensors, digital networking and intelligent data analysis enable predictive maintenance of all types of machines and devices.

Unlike preventive maintenance, the maintenance intervals are not defined in advance, but are determined and optimized based on the sensor data. Predictive maintenance therefore has advantages over preventive or reactive maintenance approaches. Complete system failures can be reduced by constantly analyzing the data and monitoring the actual condition of a machine.

During predictive machine maintenance, an enormous amount of data is recorded by the sensors on the individual machines and machine parts. In addition, information about environmental factors such as outside temperature or humidity is collected.

Dealing with these huge amounts of data (big data) is a major challenge. However, the larger these data volumes and the more sophisticated and structured their processing and analysis, the more targeted strategic decisions can be made with the insights gained.

Ideally, intelligent systems are used for this purpose, which also integrate additional information and data. This is of great benefit if a component needs to be replaced due to deviations from standards in the recorded sensor data. With the right tools, information about the manufacturer, supplier, costs, quality, expert opinions and contact persons can be provided very quickly.

Help from artificial intelligence
Big data analytics systems based on artificial intelligence, so-called insight engines, are ideal for corresponding use cases. These are the logical continuation of enterprise search applications, which now not only track down data but also help to analyze it. They also combine innovative technologies to make the provision of information resource-efficient. They help with the qualitative processing of big data and thus create a consolidation and profitable use of company data. Insight engines are self-learning and can therefore constantly expand their knowledge. They analyse context, for example user behaviour, and categorize the information based on its relevance. This provides every employee with a personalized overview of all business processes, customer and supplier relationships, departments, responsibilities, etc., tailored to their specific needs.

By linking maintenance logs, plans, technical documentation, personal experiences of experts and much more information, these solutions can provide the user with an overall picture, a so-called 360-degree view of the individual components or machines.

The Insight Engine understands human language. Natural Language Processing (NLP) and Natural Language Question Answering (NLQA) ensure that text content is correctly understood and interpreted so that the user's needs can be precisely determined. The required data is enriched with context-specific additional information and displayed as hits. In this way, users receive accurate and targeted answers to their questions instead of endless lists of results.

Even though insight engines enable the uncomplicated and rapid provision of all company data, security and data protection play an important role. Sensitive company data can only be viewed with specific authorizations. This means that each employee receives a separate view of the company's knowledge according to their authorizations. Access rights are checked for each individual query, so changes of position or department are not a problem.

Insight Engine solutions are often supplied as an appliance, i.e. coordinated hardware and search software. Integration is simplified by so-called connectors, which enable the rapid connection of different data sources such as network drives, Microsoft SharePoint and a variety of ECM systems.

For predictive maintenance, insight engines offer a good option for linking sensor data with additional information about spare parts lists, maintenance logs, plans, manufacturers, suppliers, expert opinions, etc. A 360-degree view based on this creates an optimal decision-making basis for maintenance planning and not only allows business failures to be avoided, but also processes to be accelerated and work to be carried out more efficiently.

This allows the optimum maintenance time to be identified by detecting faults or defects before they occur. Ultimately, the aim is to reduce production downtime and cut costs. This is only possible through the comprehensive networking of data from individual machines or their components with applications, departments and any business units. Only then can relevant information be retrieved by users at the right time and in the right context.