Studio manager interviewed about new Aware Systems for better onboarding
In an interview, studio manager Dr. Benedikt Gollan talks about the current development of a prototype for better training in the training phase on complex machines.
Personal:
During his studies of electrical engineering and information technology at the Technical University of Munich, Benedikt Gollan focused on feature recognition and machine learning in the application areas of computer vision and music information retrieval. In recent years he has been intensively engaged in the analysis and interpretation of human attention as input for so-called “attention sensitive systems”. In cooperation with Prof. Ferscha, he completed his doctoral thesis “Sensor-based Online Assessment of Human Attention” at the Johannes Kepler University in February 2018 and is now taking over the operational management of the Research Studio Pervasive Computing Applications.
How did it come about that you became fascinated by Pervasive Computing Applications?
I am fascinated by the vision underlying the concept of Pervasive Computing, namely to design technology so intuitively and integrate it to such an extent that it is no longer recognizable as such. Even as a student, my focus was on the interface between man and machine, and the approaches we are pursuing here at Studio Pervasive Computing, thought through and brought to completion, represent a fundamental redesign of man-machine communication. The future is (apparently) no longer having to deal with technology, but it will surround us invisibly and react to us almost as if by magic.
What new prototypes are you and your team currently building?
We are currently working on interactive systems that are able to recognize not only the activities but also the cognitive states of people and thus enable a natural, human-like interaction. Specifically, we are working on, for example, industrial assistance systems that can detect problem situations through sensor technology and modelling and only offer assistance in case of problems. The goal is a restrained, respectful technology solution that provides exactly the right assistance at the right moment. Because, nothing is more unpleasant than an unsolicited interference by a technical system.
What will this new prototype be able to do?
As the result of a research project lasting several years, this summer we will be installing a training system for teaching new employees at our project partner Fischer Sports.
To what extent will this new prototype be superior to existing products on the market?
Up to now, it has been very complicated to provide technological support for complex work processes. Our system combines behavioral analysis, interpretation of cognitive indicators, and monitoring of the product to model work progress and quality of work. This requires a sophisticated combination of different sensors, real-time data processing as well as modeling of the worker based on academically established cognitive models. Such systems have not yet entered industrial applications. Therefore we are especially proud to be ahead of the times with our partners.
What problems does the new prototype solve?
The developed technical approach aims at several problems:
1. training of personnel in acute skill shortages
2. quality management in industrial production for semi-automated processes
3. process optimization and from this also product development through comprehensive data analysis and
4. sustainable, standardisable knowledge management in the company.
What is the most important component of this new prototype?
The most important component of the training system is the underlying modular Attention-Aware Framework which is currently being developed in the studio. This allows a quick adaptation to new application scenarios and thus adaptable, individualized assistance solutions for industrial partners.
Why does this new prototype represent an added value for the users?
The methods developed enable both the automatic documentation of quality management and assistance in various semi-automated areas in order to increase the process and product quality in production.
Key Publications:
1. Gollan, B., & Ferscha, A. (2016). Modeling Pupil Dilation as Online Input for Estimation of Cognitive Load in non-laboratory Attention-Aware Systems. In COGNITIVE 2016-The Eighth International Conference on Advanced Cognitive Technologies and Applications.
2. M. Haslgrübler*, P. Fritz, B. Gollan, A. Ferscha Getting Through – Modality Selection in a Multi-Sensor-Actuator Industrial IoT Environment Proceedings of the 7th International Conference on the Internet of Things, ACM, 8 pages, DOI: 10.1145/3131542.3131561, October 2017.
3. Gollan, B., Wally, B., & Ferscha, A. (2011). Automatic human attention estimation in an interactive system based on behaviour analysis. Proc. EPIA 2011.
4. M. Haslgrübler, M. Murauer, A. Ferscha Gazor: A gaze aware Industrial IoT-based Instructor Proceedings of the 7th International Conference on the Internet of Things, ACM, 2 pages, DOI: 10.1145/3131542.3140266, October 2017.
5. Gollan, B., & Ferscha, A. (2016). SEEV-effort-is it enough to model human attentional behavior in public display settings. Future Computing.
