Binscope, a project for Gemeente Amsterdam. On this project I worked with three other students. I had the role of an interaction designer and concepter. I was the person that connected the question of Gemeente Amsterdam, the needs and wants of the target group and translated this into a meaningful concept with logic interaction. Also I was responsible for the documentation and presentations.
Gemeente Amsterdam had a complex problem for us: we need to economise money and there is a lot of data we are not using. Can you help us doing something useful with the data so we can save money?
The first thing we had to do is understanding the dataset. From Gemeente Amsterdam we received data on the disposal of the waste bins in Amsterdam and we had been asked to visualise that data so that we get more insight into the process. From each trash can we know which route it lay, where it was exactly, when and how often it was emptied and how full it was when it was emptied.
The problem we saw in the dataset is that Gemeente Amsterdam spends unnecessarily a lot of time and money on emptying trash cans that are just 10 or 20 percent full. Sometimes the trash can even overflow.
For this problem, we have developed an application that indicates how full a bin is and when it needs to be emptied: Binscope. Sensors are way too expensive to put in all the trash cans in Amsterdam and Gemeente Amsterdam has to save money. That is why we have devised an algorithm that uses data that is always reported when emptying trash can. Thus, an accurate prediction can be made of the fullness of a bin. In addition, the algorithm is self-learning because it analyses and anticipates previous predictions.
We had to present our concept and prototype at school for the client, and the client invited us to the office at Gemeente Amsterdam to give a presentation in front of other employees of Gemeente Amsterdam, because he liked the solution that much.
The complexity of this project was very high, especially because of the complex datasets. But that was the challenge: understanding the dataset, what is happening and how can we solve their problem, translating this into a concept that benefits the user and the client, and using the data to make everything smarter and better.
We have developed a prototype that gives an idea of how it would work. No login details are required. Unfortunately, there is still a bug, but by clicking on the logo you will not notice anything. The routes 22 and 23 are the only active routes at the moment. This is our prototype.