PROASSIST4LIFE

Fraunhofer Institute for Experimental Software Engineering IESE

ProAssist4Life - Proactive Assistance for critical situations - Emergency detection for senior citizens

The Situation

According to current estimates, one out of three people over the age of 65 suffers a serious fall once a year; for people over the age of 80, it is almost one out of two. Many accidents happen in private homes during everyday activities, often at night. It frequently takes hours until help arrives for the affected person. Even home emergency call systems only offer limited support, since an elderly person is often not capable of issuing an emergency alert because he or she is injured or disoriented, or is simply not carrying the push button. If helplessness is detected late, the effects can be very serious. Epidemiological studies show that following a fall and subsequent hospitalization, elderly patients are much more likely not to be able to continue living on their own, but must rather be released to a nursing facility.

Help shall be provided by an intelligent system that recognizes such emergencies automatically and reacts to them. Although promising approaches exist in research environments regarding the automatic detection of helplessness at home, these have not yet been realized in concrete products so far and are thus not available on the market – despite great demand. This supply gap can be closed effectively with this project. The project ProAssist4Life is therefore about developing an integrated software and hardware solution for home environments that can be used to detect situations of helplessness in a cost-efficient, anticipatory, and unobtrusive way, and it is about providing adequate assistance.

The ProAssist4Life Solution

The partners are working on developing an unobtrusive system that will permanently “accompany” elderly people in their own home or in a senior citizens’ home. Multi-sensor nodes mounted to the ceiling of the rooms capture the resident’s movement patterns. The data are then transmitted wirelessly to a computer. Software is used to document the resident’s activities of daily living, and thus continually learns what his “normal behavior” is. The evaluation program permanently compares the current activity of the resident with the model. This is how it recognizes deviant situations that might indicate that the person has suffered a fall, is lying motionless on the floor, and is in a situation of helplessness. If the elderly person does not react to the contact made by the system, the software will send a text message to inform a trusted person, such as a family member or caregiver.

The ProAssist4Life solution comprises:

  • A novel, cost-efficient multi-sensor node, which can be used to capture the activities and conditions in the home environment and on the part of the user;
  • An integrated framework for personalized situation recognition from the home environment, which can be used to detect and anticipate stereotypical assistance situations automatically and which can also be used to plan and provide assistance that is appropriate for the respective situation;
  • An interaction module that allows intuitive integration of the resident and his/her social network as well as professional assistance services into the assistance process, and
  • A compact middleware, which allows using the software components on different hardware platforms, such as set-top boxes.

“Initially, our solution is intended as a supplement to conventional home emergency call systems and shall improve the feeling of safety of users and family members. In the long run, both systems shall be integrated. “ (Prof. Dr. med. Christian Madler, Medical Director, Institute for Anesthesiology and Emergency Medicine I, Westpfalz-Klinikum Kaiserslautern)

The project PROASSIST4LIFE was jointly conducted with funds of the German Ministry of Education and Research (BMBF) (funded by the BMBF funding program “KMU-Innovationsoffensive Informations- und Kommunikationstechnologie (IKT) – Softwaresysteme und Wissensverarbeitung”) together with the collaboration partners