MaDCo - Integrate Machine Learning as a digitalization component in companies processes and information systems

Based on the convincing results and knowledge acquired through our previous research and implementation projects like SoLiDA (advanced machine learning), ALDAAL (improved training for machine learning), AMICAL (Active learning), MSID (OFEN), Swisstube (InnoSuisse), Buehler (mandates and patents), this project wants to go deeper in helping the digitalization effort of the Swiss companies.

The intended strategy will be to implement machine learning in existing company processes, with at first a digitalization step and then, introduce a further automatization using machine learning.

To cover different aspects of the digitalization of processes we intent to apply our research to different types of companies and different level of processes like industrial production processes (with Swisstube), machine aided human resources processes (with IpKeys) and  software automated planification processes for agriculture (with Aero41).

The main intended goal will be to establish a generalized methodology that could be applicable to many different digitalized processes for very diverse companies. The forecasted steps are:

1. Identification of to the be digitalized processes for 3 companies, collaboration with the process owner using the same methodology.
2. Implementation and performances analysis of ML in these processes using human evaluation of the quality.
3. Smart analysis of the implemented processes to establish common recipes based on the same methodology.
4. Definition of the common standardized process.
5. Evaluation of the standardized process on the 3 use cases of the project.

A joint effort with existing Innosuisse projects, a doctoral thesis and publications will also support the research and
implementation of the project.