CogLabs
"The CogLabs project is a shining example of how we can inspire and empower the next generation to embrace the transformative power of AI and robotics."
The project
CogLabs (a UNESCO project) aims to increase awareness and understanding of AI by making machine learning approachable, rewarding and most of all fun.
After successfully designing and building the proof of concept platform for UNESCO, Google and RUSH embarked on phase 2 of this world-first intuitive and kid-friendly on-device ML coding language to scale to markets around the world.
We focused on combining Scratch and Teachable Machine into a single unified platform, simplifying both visual programming and machine learning concepts while expanding on Google's visual design language.
Expertise
Experience Auditing
Feasibility Validation
Research
UI Design
UX Design
Accessibility
Machine Learning
Frontend Development
Backend Development
Quality Assurance & Testing
35
Countries participating in CogLabs workshops
200
University of Nairobi students in CogLabs workshop
>10k
Children learning about AI, ML and robotics
The PoC product combined two separate products (Teachable Machines & Scratch) and required a laptop to code, a wifi connection to push the code and a mobile to then run the ML projects. In order to scale, we needed to remove the need for a laptop and the connectivity step between the laptop and mobile device.
We first documented and refined Scratch's extensive code blocks into a carefully curated set optimised for robotics control. The development team then created an intuitive drag-and-drop interface with colour-coded blocks and ghost block previews that show users exactly where and how pieces fit together.
For machine learning, we stripped away complex options from Teachable Machine, focusing solely on straightforward image and audio uploads to make the interface more mobile-friendly.
Regular planning and collaboration between design and engineering teams across all partner organisations ensured tight alignment on work, with a quick turnaround for review outputs and testing with user groups.
"The CogLabs app's intuitive design and engaging features make it an excellent tool for introducing children to the exciting world of robotics and machine learning. We look forward to seeing its impact in classrooms worldwide."
Design principles
The app design incorporates simplicity, progressive disclosure, and user-friendly features to introduce robotics and coding to children effectively. The mobile-first interface uses colour-coded blocks for different functions, with an intuitive drag-and-drop system that previews where blocks can be placed. Machine learning concepts are made approachable through simplified image and audio training interfaces. The modular/maker space design assets, primary colours and simple shapes all make the platform approachable and engaging.
User testing & iteration
Regular testing with children and teachers in workshops provided valuable feedback for iterative improvements. Testing involved multiple families and their children, allowing direct observation of how young users interact with both the physical robot and mobile interface. Key success measures are to remove any hand-holding for easy self-directed learning, generate high effectiveness and enjoyment, and deliver an experience users want to return to. The team's collaboration and focus on user experience is critical to the project's positive global impact.
Let's make better happen.
