Psychology / Business Psychology Study Programme|PFH Göttingen
Request Free Information

GlycoRec - Interactive bio-life-logging for a better treatment of diabetes

The research project "GlycoRec" is an adaptive, learning system (also called interactive bio-life-logging), which aims to help diabetics in everyday life by making complex decisions and implementing a healthy lifestyle. GlycoRec was funded by the Federal Ministry of Education and Research (BMBF) with 1.4 million euros and was coordinated by PFH Private Hochschule Göttingen, department of business psychology. Further project partners are the company Emperra E-Health Technologies from Potsdam, the L3S research center at Leibniz University Hannover, the Ostbayerische Technische Hochschule Amberg-Weiden and the German Diabetes Center (DDZ). The project was successfully completed at the end of March 2018.


Partners of GlycoRec (f.l.t.r.): Arne Sonar, Diana Scharf, Prof. Dr.-Ing. Dominikus Heckmann (OTH Amberg-Weiden), Ilonka Wolpert (Emperra), Dr. Olaf Spörkel (DDZ), Tamara Graf (OTH), Jennifer Sabernak (Emperra), Dr. Eelco Herder (L3S), Prof. Dr. Karsten Müssig (DDZ), Matthias Selisky (PFH), Laura Dauben (DDZ), Jan Raacke, Isabel Teichman und Prof. Dr. Stephan Weibelzahl (PFH). Not in the picture are Dr. med. Janko Schildt, Dr. med. Markus Bentrup (beide Emperra), Dr. theol. Bernhard Bleyer (OTH) and Tu Ngoc Nguyen (L3S).


Diabetes is the world's most common chronic metabolic disease. About 8% of the German population suffers from diabetes mellitus and a further increase is expected. Diabetes is already one of the most frequent causes for consultations in general practice. In the foreseeable future, this need might no longer be met due to rapidly increasing cases. At the same time, it is important for patients to behave properly in order to avoid complications. The age group 60+ is most often affected; one in four 80-year-olds in Germany suffers from type 2 diabetes.
Patients suffering from diabetes have to make many decisions every day, such as: Can I eat this? Which product should I buy? Am I hypoglycemic? What can I do to lower the blood sugar level again? How much insulin do I have to take? Especially older patients and those who have been diagnosed recently are quickly overwhelmed.

Goals and Methodology

The aim of the GlycoRec project was to develop a tele-medical system, which provides an extensible, integrated infrastructure of sensor technology, modeling and patient interaction, which in turn allows tailored support for diabetic patients in everyday life.

Live Healthy

Motivate patients to lead a healthy lifestyle (balanced diet, regular exercise).

Write a Diary

Provide patients with centralized documentation support to facilitate a simplified management of a complete diabetic diary.


Inform patients about diabetes and help them to become experts in how their body responds to diabetes and their treatment.

Stabilize Blood Sugar

Help patients to keep their blood sugar levels stable. The GlycoRec system calculates a blood sugar prognosis based on the acquired data. The system learns to tailor the blood glucose model to the individual patient.<p/&g

Over a period of three years, the project GlycoRec investigated how patient support can be improved in everyday life. In line with the objectives of the BMBF program "Adaptive Learning Systems", a learning system was developed that interprets the needs and actions of patients in their treatment context and adapts to the peculiarities of the users. Special consideration was put on emphasizing the autonomy of the user. The development took into account ethical, legal and social science (ELSI) aspects identified during the project. Diabetes patients have been involved in the development process from the start, through numerous interviews, surveys, user tests and focus groups.

GlycoRec broke new grounds in the development of innovative methods and systems in several ways:

Linking of medical data and context data for everyday decisions

Medical expert systems and telemedicine measures are usually limited to purely physiological data. Through bio-life-logging, e.g. the continuous collection and analysis of data describing the life situation of the patients, GlycoRec creates fundamentally new possibilities to respond to the individual situation of patients and to jointly develop solutions and decisions in interactive counseling dialogues.

Multifaceted user model for higher fault tolerance

The use of more extensive sensor technology allows for a much more faceted and complete modeling of the user and the environment as “contextual events” compared to most adaptive systems. In particular, the continuous measurement over longer periods offers new possibilities of compensation, error robustness and error tolerance in the user model.

Individualized interactive feedback to patients

The modeling of patients and the environment offers for the first time the possibility of individualized feedback to the patient and support in areas of life that cannot be covered by diabetes counselors and physicians.

Use of user modeling methods in the medical field

The long-term user modeling of diabetes patients represents a novel field of application of established methods. The comprehension of the context and individualized counseling promises great added value in diabetes therapy and the quality of life of patients.

Consideration of ethical, legal and social aspects in development

GlycoRec acts as a companion instead of being therapeutically active. Recommendations and instructions are deliberately restrained.

The aim of this approach is to promote the level of attention and sensitivity in dealing with the GlycoRec system on the part of the users by a certain amount of self-responsibility.

Cross-platform interactions for situation-specific adaptation

GlycoRec integrates seamlessly into the daily lives of patients. This includes the fact that the networked sensor technology runs in the background and can interact with GlycoRec on a daily basis via three different platforms. The Smart Watch can be worn anytime. Smartphone and Smart TV are especially intuitive and allow a dialogue with elderly patients in particular.

Concept of GlycoRec

The project GlycoRec investigates how people with diabetes can be better supported in everyday life. Through continuous collection, storage, processing and analysis of physiological and environmental data, individual user and context models are generated, which allow to develop much more accurate forecasts and individual recommendations for the patient. GlycoRec provides an extensible, integrated infrastructure of sensing, modeling and patient interaction.

Glycorec app

Three user interfaces on the smartphone, smart watch and smart TV interaction platforms help users to solve diabetes treatment problems.
The different data streams are evaluated and analyzed to identify individual patterns in behavior and physical response, deviation from normal conditions and specific events, and to predict future blood sugar levels.

Keep Diabetes Diary

GlycoRec alleviates this by automatically recording patient data: Blood glucose values are transferred from the meter to the GlyoRec system, as is the case with the insulin pen administrations. A pedometer records the walking movements. Carbohydrate levels can be easily entered via the smartphone using an extensive meal database.

Keep Blood Sugar Stable

A blood glucose modeling system uses the collected data to calculate the trend of blood glucose levels. This is visualized as a prediction in the blood glucose chart. The patient receives personalized and context-sensitive recommendations as well as warnings to counter emergencies before they arrive. The blood glucose prediction adapts to the user data over time and becomes more and more accurate.

Plan Nutrition

An integrated meal database helps to estimate carbohydrates and calculate the amount of insulin. The database contains data from the German Bundeslebensmittelschlüssel (BLS), average recipes and meals included by users. For unusual insulin inputs, the system responds and issues a notification to provide feedback to the user.

Optimize Therapy

The data is processed in clear charts, so that connections can be easily recognized. The insight into current data and trends is possible on a smartphone or a Smart TV system. A "Tips & Info" feature provides the patient with additional information about relevant topics (such as nutrition, blood sugar effects, diabetes).


In the next step, Emperra GmbH will continue to drive the project forward to market readiness. With the transfer to the medium-sized company, we meet the requirements of the BMBF as a sponsor. The ministry had set the feasibility and commercial usability as goals. The results of the GlycoRec project are considered and integrated for the ESYSTA system of Emperra GmbH.
The ESYSTA App is available free of charge and can be downloaded from the iOS App Store and Goolge Play. Activities are currently being carried out to integrate components from the GlycoRec project into the ESYSTA system. This provides the opportunity to present well-founded knowledge in a compact form, partly tailored to the patient and tested by medical staff. In the future, therapy and knowledge will be bundled in the system, thus facilitating entry and handling in everyday life.

Project Partners

Glycorec partners

The consortium of the GlycoRec project consisted of universities and research institutes with background in user modeling (PFH, OTH, LUH), an interdisciplinary research institution that combines molecular and cell biological basic research with clinical and epidemiological research approaches in the field of diabetes (DDZ), as well as a research and development-driven medical technology companies with highly innovative approaches and solutions in e-health technologies and telediabetology (Emperra).


Weibelzahl, S., Herder, E., Rokicki, M., Heckmann, D., Müssig, K. & Schildt, J., (2015). Personalized Advice and Feedback for Diabetes Patients. In: Weisbecker, A., Burmester, M. & Schmidt, A. (Eds.), Mensch und Computer 2015 – Workshopband. Berlin: De Gruyter Oldenbourg. (pp. 305-312).

Weibelzahl, S., Heckmann, D., Herder, E., Müssig, K. & Schildt, J. (2015). Adaptive Recommendations for Patients with Diabetes. In: A. Cristea, J. Masthoff, A. Said & N. Tintarev. Extended Proceedings of the 23rd Conference on User Modeling, Adaptation, and Personalization (UMAP 2015), Dublin, Ireland, June 29 - July 3, 2015.

Rokicki, M., Herder, E. & Demidova, E. What's On My Plate: Towards Recommending Recipe Variations for Diabetes Patients. Extended Proceedings of the 23rd Conference on User Modeling, Adaptation, and Personalization (UMAP 2015), Dublin, Ireland, June 29 - July 3, 2015.

Markus Rokicki, Eelco Herder, Tomasz Kusmierczyk and Christoph Trattner. Plate and Prejudice: Gender Differences in Online Recipes. Proc. UMAP 2016.

Hao Cheng, Markus Rokicki, Eelco Herder. “The  influence of city size on dietary choices and food recommendation.” Proc. UMAP 2017.

Hao Cheng, Markus Rokicki, Eelco Herder. “The  influence of city size on dietary choices.” Adj. Proc. UMAP 2017.

Christoph Trattner, Markus Rokicki, Eelco Herder. “On the Relations Between Cooking Interests, Hobbies and Nutritional Values of Online Recipes: Implications for Health-Aware Recipe Recommender Systems.” Adj. Proc. UMAP 2017.

Markus Rokicki, Eelco Herder, Christoph Trattner. “How Editorial, Temporal and social Biases Affect Online Food Popularity and Appreciation.” Proc. ICWSM 2017.

Sonar, A. / Bleyer. B. / Heckmann, D. (2017): Zur Synergie von reflexiver Technikbewertung und E(L)SA-Begleitforschung - Eine Bewertungstheorie sozio-technischer Systemgefüge im Rahmen der Digitalisierung.

In: Seffer, K. / Kinateder, E. (Eds.): Bavarian Journal of Applied Science. No.3, Dez. 2017/Jan. 2018, S. 250-262.

Markus Rokicki, Eelco Herder, Christoph Trattner. “The Impact of Recipe Features, Social Cues and Demographics on Estimating the Healthiness of Online Recipes.“ Proc. ICWSM 2018.

List, S. (2018). Möglichkeiten der Steigerung der Adhärenz im Diabetesmanagement bei Erwachsenen am Beispiel eines interaktiven app-basierten Interventionssystems. Unveröffentlichte Projektarbeit, PFH Göttingen

Tu Ngoc Nguyen, Markus Rokicki. "On the Predictability of non-CGM Diabetes Data for Personalized Recommendation." In Proceedings of CIKM 2018 Workshops.


Brand Eins


MEDICA Magazin

Diabetesinformationsdienst München

Der Fuss - Fachzeitschrift für Podologie und Fußpflege

Orthopädie Schuhtechnik

Diabetologie Online


Informationsdienst Wissenschaft

Göttinger Tageblatt

Hessische/Niedersächsische Allgemeine (HNA)