We have begun the initial prototyping of the project. The first challenge was how to connect the clothing. It quickly became apparent that we needed a passive technology that did not require a power supply. It also needed to be washable. RFID technology meets these needs and we have considered low and high power options which offer different strengths in terms of usability, range and cost. Another advantage of RFID is that it can be powered by a Raspberry Pi, offering both low cost and a discreet size. The second area of technical development is that of the software, which adds complexity with the need to create a suitable algorithm that can effectively manage clothing usage. This software will be cloud-based and our aim is to open source it, along with the hardware specifications to foster future collaborative development.
The core technical process works as follows:
- The user tags the garments and scans them into the wardrobe adding a name, photo and usage (based on the weather and likely frequency)
- As clothes are taken out of the wardrobe or drawer, an RFID reader will identify that they have been used
- The database and algorithm will combine with open weather data to create a set of reminders to the user, for example, a raincoat might send the following message: ‘It looks like rain today. Time to wear me? #raincoat’.
- Usage reminders will be messaged to the user, and should they wish, it can also be posted on Twitter – each item will generate its own hashtag
- Clothes not worn for a period, based on the user-identified frequency, will email a charity clothing organisation (for example @fertha) who will send a mailing envelope to return the item
- The algorithm will be a learning one. Over time, it will try to understand the user’s usage patterns and make relevant recommendations taking into account additional factors such as the season, the day’s weather and when an item was last worn
- Users can view a history within the app of which items they have worn and how often. They will also be able to select items to create ‘looks’ that can be shared by email or within the app