Africa, Leading: TrAfrica

If you live in Silicon Valley, and commute to downtown San Francisco, your trip typically takes 1.5 hours, regardless of whether you try a combination of car + train + CalTrain, car + BART, or just drive yourself the whole way (and pay the equivalent of a student loan for parking). Or maybe you sign up with a startup like Luxe that asks you to hand over your keys to a total stranger and cross your fingers that they bring your car back at the end of the day. Whatever your approach, you’re aware of a glaring lack of truly effective, efficient solutions.

It could be worse…

If you live in coastal Nigeria in an area called Surulere and if you need to commute to nearby Lagos Island, your trip home can take 4 hours or more — to cover a distance of just 8 miles.

lagos<em>nigeria</em>traffic_largeA native Nigerian, Tochi Onyenokwe is currently an intern at the Spark Technology Center (STC) and a student at Harvard University: “It’s a bad situation. People working at a bank, for example, will stay at work an extra 5 hours, until 10 at night, to avoid traffic. Then you get home at 1 am, and you have to start again at 4 or 5 am just a few hours sleep.”

And the drive is rarely peaceful: tolls, the din of honking, and hours sitting still in traffic — or worse, standing in the African heat on a bus with no air conditioning

If you think the Google buses whipping by while you’re waiting for San Francisco public transit are irritating, try watching as US Consul workers are escorted in private cars and ferries to and from the Consulate gates in minutes — further extending your 4-hour, twice-daily trek.

Lagos Island is five square miles of sky scrapers and expensive residential real estate centered around a golf course and country club. Hundreds of thousands of people commute to and from the island each day, jammed onto just four routes: three toll roads and a handful of inefficient water ferry services.

While the Nigerian government struggles to improve the infrastructure, Tochi and three other IBM interns are building a Apache Spark™ application to help. They’re part of a group of twenty interns in IBM’s LEADing to Africa program, currently gathered at the STC in conference rooms that look out over the San Francisco Bay Bridge.

Tochi and his teammates (Onyenokwe, Adele Olisa, Larry Boateng Asante, and Janice Darling) hail from Nigeria, Ghana, and Jamaica — though they’re currently enrolled in elite U.S. universities. They and their fellow interns were chosen from a highly competitive field for a chance to leverage Apache Spark technology to address issues confronting modern Africa.

TrAfrica_2Left: Adewale Olisa – Rochester Institute of Technology, center: Tochi Onyenokwe – Harvard University, right: Larry Boateng Asante, Grinnell College.

And they hit the ground running. On Day 1 of development, the team had already started prototyping an app to predict traffic patterns:

“TrAfrica” will stream data – from Twitter, from users, and from the all-important toll booths at the choke point of the arterial roads, and deliver the data in a way that helps drivers choose when and how to get home. It’ll be like the radio traffic updates pouring out of just about every car window—except TrAfrica will be visualized, customizable, and streaming live.”


Because the supply of data will be limited until a critical mass of users is sending information back to the system, the team is writing algorithms to predict traffic patterns based on historical data:

“Our primary data source will be aggregated crowd-sourced traffic information from our users,” said Larry Boateng Asante, a student at Grinell College. “We’ll also collect live-stream traffic data from traffic agencies and toll booths. To model our data sets we’ll use New York City traffic data from We’re also scraping relevant information from twitter feeds, and mapping it to the NYC data sets so we have statistical data to support our predictions.”

Design: Making data available to everyone

The goal is an intuitive, easy-to-use Android app that allows commuters to input any street and find the best route to it, or to input a current location and get information about the surrounding traffic:

“We’re using Android Studio—our preferred IDE for Android Development, and writing code in Java using the Android SDK and other libraries like the Google Maps API. The maps accessed via the Google Map API will give us access to the queried streets. We then color-code these streets based on traffic flow and show the traffic density visualizations using custom markers.”

Stay tuned for visuals of the alpha app.

The team predicts that Lagos commuters could shave an hour from their commute times — an hour they could spend eating dinner with family, working, or sleeping.

TrAfrica_3Larry Boateng Asante, Grinnell College, right: Janice Darling, City College of New York.

TrAfrica is solving a problem in Africa, but it’s designed from the ground up to to work in any city in the world.

“We’re structuring our analytics model so we can handle data from any city and make a near accurate prediction using the statistical models we build off of New York City traffic data,” reports Larry.

IBM designers from the STC will work with the TrAfrica team as their app gets off the ground, and ultimately Lagos commuters could be treated to a low-cost solution that gives them a degree of real control over an endlessly frustrating part of daily life.

And who knows? San Francisco could be next.

For more information about LEADing to Africa and the Spark Technology Center in San Francisco, contact Leon Katsnelson, Twitter: @katsnelson


You Might Also Enjoy

Kevin Bates
Kevin Bates
9 months ago

Limit Notebook Resource Consumption by Culling Kernels

There’s no denying that data analytics is the next frontier on the computational landscape. Companies are scrambling to establish teams of data scientists to better understand their clientele and how best to evolve product solutions to the ebb and flow of today’s business ecosystem. With Apache Hadoop and Apache Spark entrenched as the analytic engine and coupled with a trial-and-error model to... Read More

Gidon Gershinsky
Gidon Gershinsky
10 months ago

How Alluxio is Accelerating Apache Spark Workloads

Alluxio is fast virtual storage for Big Data. Formerly known as Tachyon, it’s an open-source memory-centric virtual distributed storage system (yes, all that!), offering data access at memory speed and persistence to a reliable storage. This technology accelerates analytic workloads in certain scenarios, but doesn’t offer any performance benefits in other scenarios. The purpose of this blog is to... Read More