Uber looks to make map feature more efficient
A more efficient map feature is expected to help Uber reduce time of arrival of its cars and reduce any friction in passenger pick-ups and drop-offs
Bengaluru: Uber Technologies Inc. is striving to make its maps feature more efficient, which will in turn help the company reduce the time of arrival of its cars and reduce any friction in passenger pick-ups and drop-offs, said Manik Gupta, director of products, maps.
Gupta said the time of arrival, pick-ups and drop-offs are high on the priority list for the maps team, which is working on building efficient route directions to a granular level, for instance inside a housing society or a parking lot.
Uber is working with third party data providers such as companies that provide data on traffic or those who provide maps of a particular geography, combining it with the data accumulated through its own driver and consumer apps to make routing more efficient.
“Expected time of arrival (ETA) is super high priority for us. One of the biggest component of ETA inaccuracy is traffic. Because, if you don’t have a good understanding of traffic on a route, you cannot estimate the ETAs better,” Gupta said at a media briefing on Friday.
“Next, how we solve the pickup problem? In most cities, you have a lot of high rises where GPS accuracy is not very good. So how do we communicate to the driver where exactly to pick you up from? if you are in a mall, where should the pick-up happen?” Gupta said.
Gupta joined Uber in November last year from Google, where he served as director of product management, Google Maps,
Uber has been strengthening its maps team of late. In June last year, the company hired Brian McClendon, former head of Google Maps. The company is likely to invest $500 million in mapping technologies, The Financial Times reported in July this year.
According to Gupta, the extent of precision in maps, for instance accurate turn by turn route inside a parking lot, is not served by most map service providers, which has prompted Uber to work on its own mapping system.
“For instance, parking lots matter a lot to Uber as that is where a lot of pick-ups and drops happen. But that might not matter that much to many other mapping companies. Then you have the self-driving mode, where the level of detail is many times more,” he said.
“Maps will perhaps never be perfect. Maps represent the real world and the real world is not static. A lot of companies have invested tremendous amount of resources into that and so has Uber. So it is important for us to understand as a company what is changing constantly and accommodate them in our maps and make them as fresh as possible,” he added.