Why Esa Bus Stop?
Ever wondered how many minutes you use every day idling while waiting for a bus? What if that time could be used to learn more about activities that suit your current mood and are organized nearby you? Esa Bus Stop recognizes your feelings and provides you with empathic recommendations and advertisement that truly interest you at the given moment. Since the system is enabled by machine learning, it will learn more about the citizens and their needs as time goes by. Thus, the stop offers a great opportunity for advertisers to keep learning about their audience and their reactions as well as use the collected data to constantly enhance future marketing efforts. With this technology advertisers could display higher amount of targeted advertisements and therefor increase marketing income.
The team wanted to do something fun and learn in the process. The utilized technologies were not those we use in our daily lives since we just wanted to try out something new while having the chance.
What it does
The bus stop analyzes the user by it's sensors and displays targeted advertisements and other content depending on users's emotions, sex, age, time and whether user is going home or not.
How we built it
Esa-bustop is based on Microsoft Cognitive Services. It has neural networks which are taught to recognise people by their looks and give them recommendations by their mood. For example that 25 year old happy looking woman might want to try Yoga class this evening. We use bus stop cameras to take pictures of by passers, and send those pictures for analysis. Once analysed, screen shows best fit recommendations for that person.
Challenges we ran into
Figuring out what kind of content to show users in a different kinds of emotional states. A technical challenge we encountered was teaching the neuron network well enough in just 48 hours.
Accomplishments that we're proud of
We managed to deliver an MVP of an awesome service that we believe will add real value to both costumers and advertisers. However the solution can be enhanced with very potential features in the future. We also gained a lot of new knowledge from different technologies and from each other.
What we learned
We all learned loads about basics of machine learning, and we’ve also played with different kinds of new (or for some of us new) technologies such as Microsoft Cognitive Services and Azure that taught us a lot. We didn’t know all the team members beforehand, and since coming from very different disciplines, we were able to teach each other a lot from service & UX design and front & back-end development.
What's next for Esa Bus Stop
Making multiple interactive bus stops around the city which allows user tracking around the city with face recognition. After showing content to user neural network would track user's reactions on the content and make future content more targeted for the user. Also speech and gesture interactions with bus stop would be possible. This is how we could gather enough data about user's travel habits, emotions, places where he is going, what time and so on. Eventually bus stops could show highly targeted content based on gathered information. The data could also be utilized for city development purposes and to provide those kind of services and events to citizens that suit their feelings and needs.