uPlan

• 

  uPlan

• 

  Goals

• 

  Features

• 

  Technical Components

Inspiration

With the growing data-driven society, I was curious about how we can use data science for good and improve our society from a real human perspective, while also being mindful of our environmental footprint.

I grew up in a town where there was always land development, which is still a theme in many communities around Canada. I strongly believe that land development has an impact on individuals and society, and urban planning should take more consideration into environmental and human factors, so we can push society towards a brighter future.

What it does

uPlan is a software for urban and regional planners to visually model plans, while being mindful of human and environmental factors when creating designs. The software uses artificial intelligence to analyze the user's design and give suggestions on various aspects to reduce issues such as: road congestion, cellular connection dropping, and prolonged idling. These issues are important to address because they result in further issues such as: increased carbon emissions, the digital divide, and disrupted work-life balance for residents.

My main focus was to cover these three target requirements:

• Ensure all (remote) communities have access to cell towers in order to close the tech gap
• Reduce carbon emissions and environmental footprint to improve quality of air
• Reduce transportation time for commuters and pedestrians to improve quality of life

Common Use Case:

• An urban planner is sketching on uPlan to design her new city. She clicks the "check" button to call the AI to review her work. The AI will notice that the commercial and educational buildings are close together and the roads are narrow. Using knowledge it knows from data from other communities with similar/different attributes, the AI suggests that that specific area could be prone to traffic. The AI will put a warning sign over the section on the plan and log the information on the right hand panel.

How I built it

The software architecture design takes the datasets provided by Geotab and input them securely into a DataStax database deployed on Google Cloud Platform, then using Python and Jupyter Notebook to combine Apache Cassandra and Apache Spark to analyze data and create a machine learning model using that can classify good and bad road intersections and residential locations.

I used Axure for the UX/UI prototyping. For future web interface implementation, I will be using HTML, CSS, and JavaScript.

Challenges I ran into

This is my first time working on my own during a hackathon, so I needed to brainstorm, design, learn, and implement on my own. It allowed me push myself to understand multiple positions within a development team.

This is also my first time working with databases, so it was a learning curve for me. I utilized online tutorials and documentation to learn more about the concepts and implementation.

Accomplishments that I'm proud of

I am happy that I was able to design a product when looking at it from multiple perspectives: UX/UI design, front-end, back-end, and marketing. I am also proud that I was able to focus on a project while being at home during the pandemic.

What I learned

Since this project allowed me to explore multiple aspects of product design, development, and marketing, I feel a stronger appreciation for all various contributors of a product and a business. I also learned a lot about databases and machine learning and I want to continue learning and deeper my understanding by learning through online courses and throughout my studies in university.

What's next for uPlan

More data! As of right now, there is not enough data to properly train the machine learning models due to the many existing factors within a city (building types, road sizing, traffic signals, highway access) which affect these factors. With more research, I hope to work toward improving this project.

Built With

• cassandra
• cql
• datastax
• geotab
• jupyter
• particle
• python
• sql