By Bethan Halliwell, partner and MedTech sector specialist at European IP firm, Withers & Rogers.
From chatbots to combat loneliness to tracking systems that can detect subtle changes in the habits and behaviours of vulnerable people in their own homes, AI-based technologies have enormous potential to protect the health and wellbeing of elderly people.
With an ageing population and more people seeking to live autonomously in their own homes for as long as possible, AI and machine learning are increasingly being called upon to provide practical and reliable support systems. Chatbots trained to interact with elderly people who are living alone tend to attract most attention, but many other uses of AI are already widely used.
According to a technology trends report by the World Intellectual Property Office (WIPO), the number of people in the world aged 60 years and over is projected to grow from 901 million in 2015 to almost 2.1 billion by 2050. This is likely to bring a higher probability of functional limitations, which means more people will be affected by disabilities and the effects of ageing.
Of course, this brings with it a cost for society as governments and health and social care systems stretch resources to meet their needs.
A series of Government-backed pilot schemes have recently demonstrated the efficacy of a novel 4D imaging technology, Whzan Guardian, in tracking the movements and behaviours of elderly people living independently, with the aim of preventing falls.
Using sensors placed around the home, the technology is trained to detect deviations in the individual’s behaviour or changes affecting their living space, alerting carers accordingly. When trialled in several residential care homes, the technology contributed to a 66 per cent reduction in falls and a major reduction in emergency callouts after a fall had taken place.
Another increasingly popular application of AI is advanced wearables that are capable of monitoring biometric data – everything from their temperature to their blood pressure and blood oxygen levels – in real time and sending alerts to clinicians and/or carers as necessary.
Trained on the individual’s personal data as they go about their daily lives, these AI-powered wearables can predict when health events are more likely to occur and facilitate preventative and personalised medicine.
Despite the potential of AI technology to improve outcomes across the health and social care sector, chatbots for the elderly have also attracted criticism. Some health and social care workers are concerned that over-reliance on robotic systems could lead to a reduction in other forms of social interaction.
However, the latest chatbot solutions are smarter and as well as providing a useful source of information, they can be tailored to satisfy the needs and preferences of each elderly person. For example, Senior Talk allows loved ones to select a ‘persona’ for the chatbot based on their age, gender, hobbies and what they like talking about, to ensure it is a good match for the elderly person living autonomously.
Other assistive AI-based systems, such as MemPal, a voice-based memory assistant, can provide the elderly person with timely memory joggers – such as reminders to take their prescribed medication or suggestions about where they might have left a misplaced object.
Generative AI could bring many more benefits in the future, helping elderly people to live better lives, autonomously, for longer. The next stage in their evolution will need to address the problem of integration to ensure that next generation AI-based technologies can be used in a coordinated way as part of an individual’s Personal Care Plan.
Increased access to personal health data and greater use of predictive analytics could also bring benefits for wider society by improving patient care, public health and use of resources.
From an intellectual property (IP) perspective, securing patents for emerging technologies can help to leverage investment for ongoing R&D and accelerate the introduction of market-ready solutions. When seeking to patent an AI-based model, it may be possible to protect various parts of the AI system, including the core algorithm, the training methods or the end-use application.
Patent offices in the UK and Europe tend to treat AI-based innovations in the same way as ‘software’ inventions, which means they need to demonstrate a ‘technical advantage’ or serve a ‘technical purpose’ to satisfy patentability criteria.
However, some exceptions apply – for example, administrative uses and chatbots tend to be regarded as non-technical. In these instances, the best route to securing patent protection may be to protect the techniques used to train the system to perform its function, rather than the end use of the trained model.
On the other hand, if the training methods are hidden from view and innovators are confident that the solution can’t be reverse engineered, they may prefer to retain them as ‘trade secrets’.
Whatever the preferred approach, innovators of AI-based systems should always seek advice about how to optimise the commercial value of their innovations and mitigate risks at the same time. For example, if they opt to patent the end use application for their innovation, they will be required to disclose at least one plausible method of training the model to serve that function, and failing to do so could lead to objections due to a lack of ‘sufficiency’.
With so many practical and inventive applications of GenAI currently under development, innovators could secure a share of the fast-growing elderly care market. To achieve this, however, they should focus on developing user-friendly solutions that can be applied in a fully integrated and coordinated way, with the support of families and health and social care professionals alike.
Bethan Halliwell is a partner and MedTech sector specialist at European IP firm, Withers & Rogers.
