A Ph.D. student Ms. Ismail at the University of Canberra is working on a technology to track walk and detect if you are going to fall or not. Based on your gait (walking pattern), her algorithms can detect if you are deviating from standard strides.
This technology if brought to mass market can help elderly prevent fall and seek medical attention at a right time. This will help in reducing severe injuries related to fall and also prevent lack of mobility and isolation which arises after falling.
The great thing about this technology is that it uses a normal camera to detect the changes in your gait. Image recognition is used as an underlying technology.
For more information, refer the attached PDF.
Being a technologist, I wonder how this technology can be improved? Walking patterns can be logged in and based on learning algorithms one can classify a healthy walk and a prone to fall walk. Based on supervised learning, you can predict a prone to fall walk and alert the person immediately. As mentioned in the article also, these types of technology can be installed at day care homes to alert an elderly of their possible fall.