January - May 2018
Develop a design project using Machine Learning (ML) as the solution.
Tackling the Brief
The idea came from the lack of organization of storage in my school's design studio. I wondered if I could use ML to help resolve our clutter issue.
After doing some interviews with my peers, I found 2 opportunities: Firstly, it is to have more clarity of equipment in storage because there is lack of information about what is stored in the cabinets. Secondly, it is the ease of identifying shared items as some items are labelled by graduated students and it’s difficult to find out who they are.
The idea is to profile the objects by assigning a name to these items placed into the cabinet. Both object and facial recognition software uses ML concepts. Since a student’s information is paired with the stored object, it allows other students who would like to borrow the equipment to easily contact the individual. In the future, this system can also be used in other contexts where sharing takes place, for example in a hostel pantry.
Using a state transition diagram, I mapped out the functions Scanbinet needed to have before I proceeded to code it using python. I used a Raspberry Pi 2 touch screen with the Arduino board and a servo for the lock.
After coding and making the prototype, I did some user testing to gain some feedback about the usability of the product. Check out the end product in the video below!