Flocktracker at IdeaStream

Friday, April 13th was a big day for Flocktracker at the IdeaStream conference. Daniel Palencia presented the problem of common practices in data collection today, how inefficient the common methods are, and explained how these ineffective data collection practices are still globally used.

Image Courtesy of Gretchen Ertl

He exhibited Flocktracker’s possible applications, from in-field data collection, community surveying, collecting transport information, to disaster response. The benefits of Flocktracker in these examples have already been tested and shown in our pilot project for mapping transit routes in Dhaka, Bangladesh. We witnessed an audience excited by our product and we are very proud of the commitment of everyone involved. It was an amazing experience for the whole Flocktracker team to present to a great group of market leaders, investors, entrepreneurs, and other brilliant minds.

The event was nothing short of the extraordinary and we are fortunate to have made invaluable connections who believe in Flocktracker like we do. Below is our full presentation in the IdeaStream event: 

Video Courtesy of Gretchen Ertl

Technological change in Jakarta’s motorcycles for hire

As in many countries around the world, the advent of ridesharing apps has had a dramatic effect on transportation in Indonesia.  Arman Jalali and Rafael Milani Maderios, researchers at TU Berlin, and graduate student Faris Saffan were interested in the impact of these technologies on a particularly Indonesian form of transportation – ojeks, or small motorcycles for hire. Like for many informal modes of transportation, no official data exist on ojeks. So, to carry out their research, Faris travelled to Jakarta to collect data with Flocktracker in 2015.

Arman and his team sought to understand the differences between “smart ojeks” – those using a mobile ridesharing app – and traditional ojeks. To do this, they created a survey in Flocktracker and, after a one-week pilot with the assistance of Flocktracker staff, sent six data collectors into the field to randomly survey drivers and passengers near main public transportation hubs. After analyzing the data they collected, they learned that smart ojeks have a wider coverage area than traditional ojeks, drivers of smart ojeks report higher incomes, and users of smart ojeks report feeling safer and more satisfied.

Using data collected with Flocktracker, Arman’s team demonstrated that “smart ojeks,” or ojeks that use a ridesharing app, have a significantly larger coverage area than traditional ojeks.

Arman estimates that collecting the data with Flocktracker instead of pen and paper resulted in a 50% time savings. He also appreciated the ease of importing the data into other programs, in this case SPSS and GIS, for tasks such as statistical analysis and creating heatmaps of responses. Finally, he appreciated that Flocktracker’s cost-effectiveness allowed him to easily collect a high-quality, comprehensive dataset on a small research budget.

Arman and Rafael published their results in a recent article in Transportation Planning and Technology. After completing his Master’s, Faris returned to Indonesia to work at a German cooperative development agency.

Flocktracker in a nutshell

From transit planning to disaster response, decisions based on faulty data waste money and endanger lives. Flocktracker provides an efficient 21st-century solution to enable organizations and individuals to be their own data scientists and become the drivers in data-driven decisions.

Flocktracker 2018

Mapping informal transit routes in Bogotá


Bogotá’s urban footprint has expanded significantly since 1989, but its formal transit system, TransMilenio, has not. Residents of the newly urbanized areas have met their mobility needs through an informal network of van routes.

Until last summer, these newer, informal routes remained unmapped. Unlike users of TransMilenio or Boston’s MBTA, nobody could plan a trip on these van routes using Google Maps or even with a paper map. Riders simply had to know the routes to get where they were going or ask someone who did. Furthermore, the lack of transit data in these areas made it difficult for policymakers to plan for housing education, education, etc.

Vans wait at a bus stop in Bogotá. An informal transit system, they provide important connections to the city center and the TransMilenio Bus Rapid Transit system for residents in the new informal settlements on Bogotá’s periphery.

But all this changed when Eric Goldwyn, a Research Scholar at NYU’s Marron Institute and an Affiliated Faculty Fellow at NYU Shanghai, partnered with faculty and students at the University of Rosario to finally put these routes on the map. After testing a few different data collection apps in New York, the team decided to use Flocktracker for its ease of use, reliability, clean data outputs, and customer support. Working with Flocktracker’s developers, the team put together a customized tracker project and spent June through September riding the vans and collecting around thousands of GPS traces and information on nearly 200 van routes. This amount of data tested the limits of Flocktracker’s backend, but Eric was very thankful for the quick attention of Flocktracker’s expert developers, who quickly found a solution to the problem.

A group of University of Rosario students learn how to use Flocktracker to collect data on van routes.

Currently, Eric’s team is working to transform their Flocktracker data into GTFS, an open source structure for transit data that can be used in Google Maps and other software for analysis. He’s also trying to forge a relationship with the city and help them to get the insights they need out of this massive dataset, and he’s publishing an article on the relationship between informal transit, informal housing, and urban expansion. Next, he’s working on mapping transit systems for 200 cities around the world, including many with unmapped systems that will require extensive field data collection.

Left: a snapshot of all 600,000+ GPS traces that make up the van routes dataset. Right: raw data for one route, ZP 35.

For more on Eric’s data, check out this blog post from Flocktracker alumnus and data scientist Kuan Butts.