Chih-Hsuan Chen

Multimodal modelling and motion prediction in dynamic environments

Principle Supervisor
Dr.-Ing. Birgit Graf
Fraunhofer-Institut für Produktionstechnik und Automatisierung IPA

Collaboration partners:

  • Universität Hamburg
  • Fondazione Istituto Italiano Di Tecnologia

Competence Area: Embodiment


Owen Chih-Hsuan CHEN obtained the master degree in electrical engineering from Yuan Ze University, Taiwan in 2010. He studies in the field of humanoid robot balance and walking control. During his study, he has granted several awards including the First-Place winner in Asia FPGA design contest.

He has been working in the industry for 6 years after graduating from university. He worked as technical leader and systems software engineer in server industries group at Hewlett-Packard (HP) for 3.5 years, then he worked as embedded system and control system engineer at HIWIN Corp. for 2.5 years, the first industrial robotic company in Taiwan, where he has published 2 patents in the field of industrial robot applications.

Currently, he is doing research in the department of “Robot and Assistive Systems” at Fraunhofer IPA since July 2017 with Marie Curie Fellowship as an Early Stage Researcher. His research focus is on modeling dynamic environment and dynamic objects in the human environment. This allows indoor service robots to identify places where they can operate safely.

Detailed CV


Tracking the precise location of humans and dynamic obstacles allows service robots to distinguish between places where they can operate safely and others, where collisions and thus safety-critical situations are likely to occur. Therefore, in this project, vision-based environment modelling, that is 3-D point cloud generation as well as geometric mapping will be integrated with other perception modalities such as sound-source localisation and touch sensors. Furthermore, motion tracking and prediction functions will be developed based on the integrated sensor input which will be provided in real-time to higher-level, cognitive modules.

Expected Results

Methods to efficiently process and store this enhanced environment map will be designed and implemented, that are specifically suitable for real-time collision avoidance. Besides providing functionality to give fast feedback on the risk level of planned actions, directly affecting behaviour by prohibiting dangerous actions, sophisticated real-time 3-D environment maps will also form the input for higher-level, cognitive safety modules, enabling them to make faster decisions based on more precise predictions.


  • Sep. 19-20, 2017: Attended the 1st SOCRATES project workshop, Barcelona, Spain.
  • July 19-20, 2017: Attended the ROS-Industrial training (EU), Stuttgart, Germany