Human Preference-Aware Optimization System (HPAO)

Project Overview

The ready availability of a wealth of data enables machine learning-based optimization from the local factory to global value chains. Existing and envisioned uses of such AI-driven optimization focus on optimizing efficiency, profitability, or process parameters such as just-in-time production and delivery – prompting ethical concerns, especially regarding loss of autonomy in the workplace, as well as increased stress levels and health-related losses among employees. The HPAOS project, led by Tim Büthe jointly with Johannes Fottner, develops an alternative approach to AI use. It utilizes data about employee needs and preferences to develop a human preference-aware optimization system, which assigns shifts and tasks based on employees' non-invasively measured preferences, promoting their strengths, respecting individual differences, and even strengthening employee autonomy.

Project Members

• Tim Büthe
• Charlotte Unruh
• with several HfP/TUM student researchers

+ Prof. Dr. Johannes Fottner and Charlotte Haid (TUM School of Engineering & Design, Chair of Logistics)