Alzheimer task identification based on stimulus-response compatibility for digital kitchen environment

Ambient Assisted Living (AAL) is an embedded intelligent object to support people in their daily activities, especially for elderly people to live independently and monitored. AAL is based on ambient intelligence (AmI) that broaden the interaction between human beings and digital information technology in a smart environment. This research implements AmI system into smart kitchen environment in activity daily livings (ADLs) among Alzheimer’s patients. These patients suffer from declining of mental functions that lead to disruption in their daily activities. The purpose of implementing AmI is to monitor and assist the Alzheimer’s patients in their activities and behaviours in the kitchen environment. In addition, this technology is crucial to encourage them to be independent in the daily living activities especially in the kitchen while their caregivers are away. Currently, the functionalities of kitchen are not supporting and user-friendly to the Alzheimer’s patients. This is due to the lacks of technology in order to monitor and identify abnormal patient’s behaviour in the kitchen environment. To overcome the aforementioned problems, this research applies the concept of stimulusresponse compatibility (SRC) in the kitchen environment whereas the cooking guiding activities are supported by using ambient cues. Specifically, an individual’s activity identification can be identified based on Hierarchical Task Analysis (HTA) and Probability Tree Diagram. A prototype of smart kitchen counter for making a glass of juice have been built with the implementation of ambient lights. The ambient light elements (stimulus) are aimed to guide the patient's actions (response) in order to complete the task according to predetermined order. An experiment of completing a task of making a glass of juice has been tested between Alzheimer’s patient and normal people. The results are collected based on three main categories which are their acceptance towards instruction given, identification of correct equipment to be used and respond towards ambient light display. From the obtained results, both Alzheimer’s patient and normal people can clearly understand the given instruction even though Alzheimer’s patients take longer time to finish. Along the experiment, 5.47% of normal people make less mistakes in identifying the tools compared to 28.1% of Alzheimer’s patient that keep repeating picked the wrong tools. The 33.6% of participants that make errors can correct the mistake through guidance from ambient light display. Therefore, ambient light can highly influence the actions of patients as it guides the patients to the correct step if they do wrong while completing the task. As obtained through probability tree diagram analysis, Alzheimer’s patients are capable to handle only the simple action such as picking and taking utensils clearly compared to the complex actions.