EMMA Technology

Lord Darzi’s (2008) review and 10-year vision of the future Health Service, The successful implementation and delivery of Telecare requires a ‘whole systems’ approach.

Using an integral system EMMA can detect human falls with a sensitivity of 97.5% and specificity of 99.5%. It is a reliable, low power, adaptive monitoring system for the elders.

EMMA (Elder Mobile Monitoring Assistant) is a monitoring and social aid for the elders to be used in their home environment. It is composed by a wearable sensor which monitors and detects falls (typically indoor) and a web social network called EMMA family network to support the need of social contact of elders.

EMMA is aimed at assisting elders socially and physically. It is a combined solution: a sensor that monitors position and vital signals of the elder and a social website (EMMA Family: see EMMA Family Web) to connect the user to his friends, relatives and organizations. The device alerts automatically a list of predefined phone contacts. Additionally the user can also ask for general help by pressing a button.

Falling down at home is the most common accident for elders. In the UK, each year there are about 2.7 million accidents which necessitate a visit to hospital. A typical domestic fall will cause unconsciousness, hypothermia and in 72 hours a fatal death. Falls accounts for 10% acute hospital admissions, with the total cost of NHS amounting to £1.7bn.

Existing products for fall monitoring rely on an internet connection (via Wi-Fi) or on a Bluetooth connection with a palm or mobile phone device. Some others are based on a radio connection with a base station. The problem with these configurations is that if the radio link (RF link) or the Wi-Fi connection or the Bluetooth connections are lost, the device is not able to signal the event (a fall typically) to its emergency list.

This invention adds another layer of communication, the GSM unit to the wrist sensor and the base station. Therefore if the radio link between the wrist sensor and the base station is interrupted, the wrist sensor uses the GSM unit to send the alarm to an emergency list. Same idea applies to the base station which monitors the internet connection and uses the GSM unit when appropriate.

The base station has also the role of monitoring the status of every wrist sensor such as the battery level and the operating status of the onboard sensors. The wrist sensors and the base stations are connected using a mesh network (like the ZigBee standard) to reach a low power profile.

The advantage of using a meshed network becomes more relevant when one scales up the solution to more wrist sensors, say on the order of 100 units. This feature is very important in a nursing case scenario where is necessary to monitor more than 1 person.

The sensor is based on machine learning algorithms. EMMA monitors position, vital signals and tasks of the elder. EMMA learns the walking pattern of the person wearing it and so is able to discriminate about normal walking patterns and dangerous ones. In case of dangers EMMA can send emails, sms or call a predefined list of contacts. The product is composed by the sensor hardware and the software for the data analysis. At the moment the sensor requires internet connection but the final version will include a GSM module. Re-design of the electronic board; software improvement test and certification (CEE and Rohs) are necessary to become a commercial product.

How EMMA works