Many applications of cyberphysical systems rely on an integration of geospatial data and sensor data. In the engineering industry, dynamic mission planning of service technicians and locating suppliers can benefit from such integrated data. Other potential applications include intelligent parking and refueling by finding available parking spots and fuel pumps or charging spots nearby. Sensors of satellite navigation systems in cars and intelligent fuel pumps, connected charging points and industrial machinery generate terabytes of industry-relevant data every day. Combining many data sources is the most promising approach, but this is difficult. Relevant geospatial data is distributed among structured (e.g., sensors), semi-structured (e.g., OpenStreetMap) and unstructured (e.g., Twitter) data sources. Due to the significant volume and variety of data sources, innovative solutions are required for the acquisition of geospatial data, integrating them with sensor data and building intelligent services on top.
The GEISER project aims to design and implement innovative functionality for developing services for transforming, storing, integrating and processing geospatial and sensor data. Here, machine learning approaches will be applied for tasks such as computing topological relations between resources and time-efficient generation of link specifications. The resulting tools will be integrated as microservices in an open cloud-based platform.