Process Mining in building operation
Process Mining is a young, analytical discipline for recognizing, monitoring and improving real business processes (i.e. processes that have not been adopted). It extracts knowledge from event logs available in today's information systems (software used to support them). In general, process mining can be understood as a bridge between data and process science (see van der Aalst, W.: Process Mining: Data Science in Action, 2016, p. 16).
In the construction and real estate industry, complex and individual projects are generally handled. Such processes present great challenges for the applicability of process mining methods. The higher the degree of standardization of the processes, the better process mining techniques can be implemented.
Wherever process-related data is generated, process mining analyses are a good choice. Especially in building operations, there are defined processes that can be validated and optimized with the help of process data. However, since the application of process mining in this area has not yet been tested to any great extent, pilot use cases must be created using selected CAFM modules.
CAFM modules according to GEFMA 444
Locking system management
Safety and occupational safety
Help and Service Desk
Budget management and
BIM data processing
Since the still young, data-based analysis method of Process Mining is hardly used in the construction and real estate industry, the following research objectives, among others, result:
Standard implementation of process mining analysis methods in the building sector
Acquisition of know-how regarding obstacles and solutions for data preparation and exchange
Generation of knowledge regarding actual processes in building operation
Detection of unwanted sub-processes and flat necks
Determination of best practice processes
Enabling error prediction before negative effects occur
Answers to questions such as the following:
Are target processes adhered to?
Are deadlines/deadlines met?
At which point do efficiency-enhancing measures have the greatest leverage effect?