This document presents a summary of our whitepaper. The complete document is available at:


Citizens in every city now use data from smartphones & embedded satellite navigation to work around traffic. Since current services on the market do not provide real time car congestion, its difficult to predict traffic flow, demand user flow, demography effect by region, weather conditions, planned or unplanned events, street size and roads, traffic lights optimization. Traffic jams cost the US a staggering $87 billion in lost productivity per year.

Cities are growing incredibly complicated. Urban developments haven't changed that much in the latest century, but urban mobility and technology has changed dramatically.

New mobility services like electric scooters, shared bicycles and on demand car ride services like uber are mixing up with traditional mobility options such as buses and metro to create new options for citizens.

It's becoming a trend to take hybrid mobility options. citizens combine different mobility services to reach their destination. These hybrid mobility options are dynamic and vary between personal preferences of citizens, but the transportation method chosen is also dependent on seasonal factors like weather, and endemic factors such as slopes or safety.

What it does.

The main goal is provide access to safe, affordable, accessible and sustainable transport systems for all, improving road safety, notably by expanding public transport. It requires flexibility, resilience, resourcefulness, and a certain degree of creativity. With the application of current solution we aim to impact five core element: Individuals, Environment, Government, Productivity, Saving.

Overall it provide accurate recommendations prior to the travel with roads to follow and transportation type been use by the user base on all his input information including, destination, road safety, congestion, income and saving recommendations, people near by, travel history, ecological impact, constructions nearby, demography, transportation system demand etc.

How we built it

The proposed system learns as information changes, and as goals and requirements evolve. It resolves ambiguity and tolerates unpredictability. It is engineered to feed on dynamic data in real time.

Since the solution to A involves B, and the solution to B involves A. We can use a classical algorithm for solving such a problem, that is the Alternating Least Squares (ALS) algorithm. This framework has many advantages for building multidimensional time series data: 1. The framework can maintain the original data representation in the form of a matrix. 2. The framework can reduce the amount of parameters in autoregressive model.

Challenges we ran into.

On the four weeks of developing this project We spend one week reading websites and pdf about machine learning and AI, reading documentations, and watching videos about TigerGraph. The main difficulty we face was selecting the challenge to work on and build since we had many ideas to work with. But finally we took one of high impact.

Data Modeling and Data Collection. Fully Learn how TigerGraph Works. Building the Team. Translation of all the content and video tutorials from English to Spanish since we are Latin Speakers.

Accomplishments that we're proud of.

Finish the Project on time. Extensive research and knowledge earned. Demo Video. GitHub Example Written Defense. Creation of Variables and Functions to integrate on system platform.

What we learned.

Get Started with TigerGraph Tutorials. TigerGraph Machine Learning Workbench OvervieW How to create a SampleProyect on Github. How to use the developer program and library from TigerGraph. Integrate Blockchain Terms into the solution proposed. Powerful GSQL features such as dynamic arrays Using Tiger Graph with Docker

+ 263 more
Share this project: