The CSIR and Uber have signed a collaboration agreement to share expertise on the use of Uber transactional data and city transport models to help cities to plan better, including the effective evaluation of where transport infrastructure investments are most needed.

South Africa has a transport-intensive economy and requires infrastructure and services that need to serve a dynamic population that include many people who are still being excluded from optimal participation in the economy due to transport barriers. The allocation of the country’s limited resources for transport infrastructure improvement needs to be carefully considered based on the most reliable data urban planners can afford to procure.

“In transport planning, data collection is an expensive undertaking, and very susceptible to budget cuts at the expense of proper transport plans,” says Dr Mathetha Mokonyama, manager for transport systems and operations research at the CSIR.

“We have often worked with government to conduct household travel surveys to ascertain how individuals travel. City planners often rely on data collected via such household surveys. These surveys are expensive and the data are not always reliable due to the nature of questionnaires and the fact that travel behaviour changes rapidly in cities.

“This is especially so in Gauteng where there is high migration and a fast-changing built environment. The data collected last year might not be applicable today, even for strategic planning purposes.  It has therefore always been our aim to find more cost-effective ways of continuously collecting the required data,” says Mokonyama.

The CSIR and Uber’s experts have been collaborating with Uber for the past year. The data collected by Uber vehicles give an accurate log of travel times between nodes such as Sandton and the Oliver Tambo International Airport, for example, by time of day, day of week, and day in a month.

“If there are changes in the built environment, new lanes in roads, bridges or developments, one is able to immediately detect the effect on travel times on a daily basis, based on Uber transactions in an area. City planners can use this data to more effectively refine their interventions,” says Mokonyama.

Uber has launched the Uber Movement website where users can access trip data over time. Thousands of trips reveal trends that are much more accurate to city planners than costly low sample surveys with high variances. Key planning variables include departure time, arrival time, and choice of destination, which are measured retrospectively. One can also assess the effect of travel conditions such as severe weather events. 

“The CSIR helped Uber with their user interface and to package the data in such a way that is compatible with existing transport planning tools used by cities in South Africa. For now, we are working in Gauteng, but the plan is to extend the platform to other areas.”

According to Mokonyama, there are other options for transport data collection, for example mobile phones and other forms of transactions. “We have been looking at that for the last few years. In the case of mobile phones, network providers can provide data without compromising privacy. This would allow for improved matching of service designs to travel patterns,” he says.

He also believes the current public transport system does not meet the needs of the end-users, resulting in high cost of living and of doing business.

“In other countries, transport is a voting issue. It is so central to the daily lives of people that it plays a role in how voters judge political candidates. If you empower people with transport system data, it is much easier to hold the authorities accountable.”

Mokonyama will talk about smart transport solutions for South Africa as part of the CSIR’s Smart Infrastructure lecture series at 12:00 – 13:00 on 31 October 2017 in the Ulwazi room at the Knowledge Commons, CSIR Pretoria campus.