Falls detection technology developer HomeGuardian has partnered with aged care provider Villa Maria Catholic Homes (VMCH) and Deakin University’s Applied Artificial Intelligence Institute to run a three-year, $10 million project to try to predict and prevent falls.
The study will see more than 1000 devices installed in aged care facilities across Victoria in the first year, with a range of data collected from the monitoring systems as well as clinical data from nurses.
The AI-powered assistive technology device automatically detects when something is wrong such as a fall, incident or illness symptoms, and can alert caregivers within two seconds. It is already being used in people’s homes and care environments.
Along with detecting falls, the project aims to go a step further and try to predict and prevent falls.
According to HomeGuardian, using predictive indicators to reduce falls incidence would make the device the first product in the global market to provide this functionality.
The study, which has started, aims to advance fall detection by combining nursing and aged care expertise with advanced AI technology to recognise when a fall is more likely to occur and to trigger a series of interventions.
“We are trying to predict a fall before it happens rather than caregivers or nurses being notified when a fall has happened,” Deakin research team chief investigator Rena Logethetis said.
“At the same time, because we are collecting a range of data both from nursing staff and from the HomeGuardian monitoring systems, we are also looking at what are the causes of a fall as well as what can be put in place to help prevent falls.”
The team will recruit 1000 participants and collect technical data from the monitoring systems and observational studies undertaken by nurses at VMCH facilities nationwide.
Causes of falls range from slipping, tripping or stumbling hazards to medication or illness which may cause dizziness, furniture, fatigue or balance and strength issues.
Analysis of behaviours that lead to falls
HomeGuardian’s founder and co-CEO Kane Sajdak said the behavioural analysis device analysed the behaviour of humans in a certain space and if abnormal behaviour – such as falling – occurred, it alerted caregivers.
“The real power is that HomeGuardian is constantly measuring that telemetry data without compromising the privacy of the people using the product,” he told Pulse+It.
“With enough data, HomeGuardian can then analyse behaviours that lead to that adverse event – and through machine learning and secondary AI it can then start to identify the indicators that occur before a fall, and alert based on those.”
The patented device has no video or image streaming but when it determines something isn’t right, it automatically alerts up to three chosen contacts without needing to press a button, make a phone call or activate an alarm.
It monitors and understands how many people are in an area or if the individual is absent and also identifies wandering and what behaviour would be considered outside of their “normal” patterns.
HomeGuardian says if the device detects a potential incident such as a fall, the device’s status light will change from purple to orange but will also assess the environment to ensure it doesn’t create a false alarm.
Once the device’s status light turns red, the device has confirmed the incident and is ready to automatically raise an alert with emergency contacts. The resident does not need to activate anything, press duress buttons or use voice commands.
The company also produces an SOS button to use as a portable duress alarm. It charges a monitoring fee for uninterrupted incident monitoring, remote device maintenance, and updates to the artificial intelligence. Users are also given access to a secure portal to manage emergency contacts and device preferences.
It also predicts illness symptoms such as coughing or sneezing, and behaviours such as wandering, absence, changes in behaviour and decline in health.
“All of the additional functionalities will be used as some of these also increase the risk of a person falling – fatigue and illness for example are high indicators of falls risk increase,” Mr Sajdak said.
“However, the focus of the study is on falls prediction as the grant required a specific goal. All elderly people who receive a HomeGuardian as part of the study will get access to all features and benefits.”
Reducing the burden on hospitals
The total project cost and deployment is more than $10 million, of which the federal government has contributed $2.65 million through the Cooperative Research Centres Projects Grant program which supports short-term collaborative research projects for products that will solve problems for industry and deliver real outcomes.
“Once HomeGuardian can predict falls, it will potentially be able to reduce the burden on the hospital system and its over $4 billion spend on treating elderly people who present to hospital with a fall each year,” Mr Sajdak said.
“Preventing the leading cause of traumatic injuries and death among the elderly will take significant pressure off healthcare and aged care systems around the world.”
The funding will allow thousands of HomeGuardian devices to be installed across VMCH’s 13 aged care homes and provide support to monitor falls to help identify and enhance preventative measures. They may also be in common areas, not just resident rooms.
“This technology has the potential to not only help reduce hospital admissions and injuries, but also save lives,” VMCH chief information officer Maria Paz said in a statement.
“Even when falls don’t cause injury, they can trigger a loss of confidence and an ongoing fear of falling. This technology will help boost independence that is so important to maintain as people age, plus provide peace of mind for families and carers.
“We are absolutely thrilled to play a key role in such an important initiative. The grant will allow VMCH to continue to be a leading innovator in the adoption and development of technology solutions, through industry and research partnerships, to positively impact the broader sector.”
Mr Saidak said he hoped the industry would rapidly adopt it as the industry standard home monitoring and personal alarm of choice.
The study will run until December 2025.
WOW, what a waste of money, did we not learn anything from the millions of dollars that South Australia sent on these systems to realise this number of false positives from analysing “fall data does not work, especially if some one is off access to the camera.