Mzansi Move

• 

  Slide 1

• 

  Slide 2

• 

  Slide 3

• 

  Slide 4

• 

  Slide 5

• 

  Slide 6

• 

  Slide 7

• 

  Slide 8

• 

  Slide 9

• 

  Slide 10

//Inspiration Mzansi Move was inspired by a deep dive into South Africa’s minibus taxi industry, one of the most vital yet informal public transport systems in the world. Minibus taxis transport approximately 15 million commuters daily and account for the majority of public transport trips, especially for lower-income frontline workers. Despite its importance, the system is unpredictable, unsafe, and largely unsupported by modern technology.

//What it does Mzansi Move is an AI-assisted taxi coordination application that helps drivers optimize routes and pickups while preserving the existing cash-based taxi culture. The app structures journeys into “trips,” dynamically manages passenger queues, optimizes drop-off order, and enforces vehicle capacity. This allows drivers to maximize passenger volume and profits while reducing reckless driving, and gives passengers safer, more predictable journeys.

//How I built it I began by researching how the taxi industry operates in practice. I then fully wireframed the application and designed the trip logic in Canva. Using Google AI Studio and Gemini AI, I provided a detailed PDF outlining system behavior, UX flows, and AI-driven trip logic. From there, I iteratively refined the application through testing while learning how to integrate route optimization concepts using mapping APIs.

//Challenges I ran into The biggest challenge was designing AI logic that respects real taxi behavior—flexible routes, cash fares, and driver autonomy—without imposing unrealistic constraints or increasing driver risk.

//Accomplishments that I’m proud of I successfully configured the AI to follow a locally relevant trip-optimization algorithm tailored specifically for South African taxi operations, ensuring fairness, efficiency, and safety.

//What I learned I learned how to collaborate effectively with AI tools, design AI-first systems, and understand how route optimization and mapping APIs work in real-world applications.

//What’s next for Mzansi Move Next, I plan to build an ai that can connect to a Dashbaord camera to that is trained on good versus bad driving in south africa that will then share a more accurate rating for drivers within South Africa.

Built With

• css3
• google
• google-cloud-run
• google-maps
• html5
• react
• sql
• tailwind
• typescipt
• vite