When it comes to the daily activities that pose difficulties for people with dementia, getting dressed is one of the big ones. Not only can it be frustrating, but it can also be embarrassing, if they have to be helped in what is normally a private activity. It was this in mind that a new "smart" dresser was created.

Known as DRESS (Development of a Responsive Emotive Sensing System), the prototype device was developed by researchers from New York University, Arizona State University and Boston's MGH Institute of Health Professions. It consists of five drawers, with a tablet, camera and motion sensor located on top.

Before it's time for the user to get dressed, their caregiver loads each drawer with a single item of the user's clothing, in the order that they should be put on. After the caregiver leaves the room, a voice cue (recorded by them) prompts the user to open the first drawer, which simultaneously lights up.

Using image recognition and tracking systems, along with a visible barcode on that first item of clothing, DRESS then observes the user putting it on, noting what type of clothing it is, along with its location and orientation on their body. The system lets them know if they've put it on correctly, and if they have, then they're prompted to open the next drawer and put on the next barcode-tagged item, and so it goes until they're fully dressed.

DRESS also advises the user if they haven't put an item on properly, plus it monitors their stress levels via a skin conductance sensor that they wear as a bracelet. If the system detects that the user is getting frustrated, it will alert the caregiver via an app, which the caregiver can also use to remotely track the user's progress.

To test the system, 11 healthy volunteers used it to get dressed properly, plus they intentionally made mistakes such as putting on items inside-out or backwards, or not getting fully dressed. While DRESS was quite good at identifying the type and orientation of clothing, along with the person's current state of undress, it had some trouble determining when they had completed putting on each item.

The researchers plan on addressing this limitation by increasing the size of the barcodes, minimizing the folding of clothing in order to keep those barcodes visible, and optimizing the positioning of users relative to the dresser.

"With improvements identified by this study, the DRESS prototype has the potential to provide automated dressing support to assist people with dementia in maintaining their independence and privacy, while alleviating the burden on caregivers," says NYU's Dr. Winslow Burleson, lead author of a paper on the project.

That paper was published this Tuesday, in the journal JMIR Medical Informatics.

View gallery - 2 images