Prevention of dependence represents a major healthcare issue. Frailty is a syndrome determining a higher vulnerability to stressors. It is responsible for an increased risk of major negative health-related events, including disability [1-2]. Fried et al. criteria are the most commonly used by the medical community to identify frail subjects: slow gait speed, less muscle strength, exhaustion, sedentariness, and involuntary weight loss [3]. In order to encourage sedentary people to practice walking, we propose to develop an economically viable technical device facilitating the follow-up of frail elderly people. In a previous paper [4], we have presented the electronic design of a smart shoe insole to quantify the human movement without constraint of positioning. This device embeds some sensors and a wireless communication. The smart insole allows measuring the gait speed and its variability, the weight variation and daily activity with minimum invasiveness. It performs measurements continuously and automatically during the walking periods, both indoor and outdoor. This wireless insole transmits information to a local database which can be consulted from a secure Internet connection in real time. This paper focuses on the algorithms that evaluate the stride length. It presents several methods for measuring the gait dynamic parameters and gives preliminary results.