Overview
Oscar is a trash can powered by machine learning that detects whether there’s organic matter present in recyclable plastics. Oscar introduces children to responsible waste disposal practices by sharing light and informative messages.
Duration
3 weeks
Focus Areas
Machine Learning
Industrial Design
Interaction Design
Why should we care?
Microplastics have been found in air, water, food and now... human blood
USA Today | March 25 2022
Our Mission
To encourage responsible waste disposal amongst children and automize waste management at educational institutes
The Solution
Welcome message
The default view expresses a message for responsible disposal practices
Ideal scenario
A rewarding experience when the child tosses an empty recyclable plastic item
Error Message
It’s crucial to educate the child if their plastics contain organic material
The Outcome
Reviewed by industrial experts
“This is a powerful medium to establish a change in mindset amongst children to get more responsible in waste disposal practices.”
Shamik Ray | Spotify | CIID
“There is potential to workout a promising monetization strategy for this product. I suggest you think of ways to target additional age groups.”
Sarah Ludwig | Design Manager at Amazon
The back story
Why a trash can focused on plastic?
Inspired by waste segregation done in certain European nations
We narrowed our focus on Plastic.
Glass
Plastic
Paper
Metal
The primary concept
Leverage Machine Learning to create the most valuable experience.
Reject trash
Con: Children may leave the trash outside the can.
Toss trash back
Con: Risk of injury if the trash is thrown back.
Auto sorting
The system will be more efficient in segregation of empty and filled plastics.
Meet Oscar
Our sentient personality from Sesame Street who thrives on plastic.
Oscar could render a fun and engaging experience for children aged five to thirteen.
Programming that gave life to Oscar
We had to learn machine learning and programming tools
Teachable machine helped us train the system to identify bottles that contain liquid.
PictoBlox was used to program what message needed to be displayed for the respective scenarios - default state, success and failure.
Other programming tools that we explored
Processing
There were possibilities to obtain source codes, however, we wanted to build the code ourselves.
Wekinator
Helped us train the machine to identify liquids but was not compatible with PictoBlox.
Tramontana
Was challenging to integrate with our code since it would often crash unexpectedly.
Introduction to industrial design
Form was envisioned to represent the fusion of nature and technology
Option 1
Con: The display won’t be visible from a distance - missed marketing opportunities.
Option 2
Con: The form appeared refined on paper, however on building the model it appeared too chunky.
Option 3
Sharp lines compliment the floating interface that can be viewed from further away.
Mycelium was selected as the primary material
It’s recycled and regrown rapidly. Additionally, studies suggest that it makes building materials fire-resistant, lighter and stronger.
Version 1
A basic massing model to share the concept and interface design.
Version 2
Criticized for not appearing as a trash can that would be used by children.
Version 3
Creates an illusion of Oscar rising from the trash can.
Trash level indicator
Intentionally indicated separately since the interface could highlight sponsored products.
A quick reference for the staff to understand when to empty the bin.
A sneak peak into the earlier wireframes
Generic messages that were later refined to be conversational with children
Future Scalability
Other types of trash - more characters
Marketing opportunities in the food and beverage industry
A web series on Netflix
Data analytics on the cloud
Reflections
Challenges faced and what I’ve learnt
Bottlenecks
Prototyping as a subject was super intimidating due to the requirement of coding and electronics - mediums that I’ve never worked with.
Often times our ideas would fail and we were forced to think of a new direction.
Learnings
Always dedicate twice the amount of time you may think is suitable for such work.
Failing early will build wisdom that will help in adapting to time constraints and limited resources.