Virtual Assistant for Assistive Living

In this digital era it is not a dream anymore to have a 24/7 assistance looking after our personal wellbeing. Virtual Assistant can play a crucial role managing the challenges of our daily life, especially for people that may need an assistance to perform a task. Self-care is a very specific domain where Virtual Assistant (VA) can provide ambient assisted living. Virtual Assistant can simplify everyday life, increases the comfort of living and the foremost objective of using Virtual Assistant is to assist people with medical conditions which will enable them to manage everyday life. VR can be proved most useful in personal care especially for people who require assistance for living due to physical impairment. The application of VA can be extended further, ie. assisting teenagers who are required to manage a house when their parents are at work; taking care of elderly people who are living on their own. The application of VA requires developing a system interface which interacts with server based Avatar through a dialogue manager and presents a dialogue to the interface in order to establish a communication between user and VA. The objective of this system is to be able to detect users’ command and their preferences through a dialogue, then, finally to execute an action on behalf of the user. Some examples of user centric actions that a VA can learn based on users’ context profile (habits, safety precautions etc) and perform accordingly are:

  • Kettle on/of, light on/off, radiator on/off
  • TV on/off, music/movie channel on/off
  • Operation on/off for other household devices

In addition to this, VA can also assist users to manage their medical conditions according to the personalized care preference. Some examples are:

  • Reminding of medication and alerting patients when medicines have been ignored
  • Alerting on internal and external contexts, i.e. humidity, rain, pollen, traffic, to manage the users’ medical condition
  • Alerting and contacting next-of-keen (stakeholders) when users’ habit/safety context is compromised, i.e. long waiting time, delay in return time, etc.
  • Assistance in shopping, in education and for web browsing.

 

In a summary, the project aims to develop a system involving Virtual Assistant which will learn from users’ habit and contextual profile (i.e. bed-time, tea-time, distance travel for a walk, waiting time for a bus), as a result it will adapt the preferences based on the users’ interaction to the Avatar.  System will adapt users’ behavior through a learning and transformation process.

Example projects:

Context aware Virtual Assistant for Asthma medical condition management. Ph.D work (ongoing).