Similar to a lot of things we take for granted these days, a couple of years ago it was unthinkable what Google Street View now offers to its users. It does an incredibly good job in capturing the world and letting us explore places all over the world while enjoying a warm cup of coffee in our comfortable sofa. So far, no other company has been successful in reproducing the content Google Street View has captured. However, a Swedish start-up, Mapillary, is trying to do things differently since its inception in September 2013. Instead of constructing a representation of the world on their own, they are building a crowdsourced competitor of Google Street View by making use of pictures taken by their community of passionate contributors.

This month, Mapillary has raised $8 million in a Series A investment round. With the additional funding, the company is able to expand its operations to San Francisco and boost its computer vision engineering capability by hiring new staff. After last years’ $1.5 million seed funding, the company has now gained extra power in order to compete with Google Street View (Butcher, 2015; Lunden, 2016).

So, how does it work?
The ultimate goal of Mapillary is to visualize the entire world. In order to achieve this, Mapillary relies on the help of its users, who actively participate in capturing parts of the globe. It’s fairly easy: walk, bike or put your smartphone in your car, make photos manually or let the app take pictures of the surroundings. Mapillary connects these photos with others’ and stitches that together into street level view. Computer vision on the server-side automatically matches and combines photos across time and across users. The resulting aggregated map can be used to explore (parts of) the world. Currently, the live counter on the homepage of the website shows that 52,616,559 photos have been taken with Mapillary, covering 1,293,850.4 kilometers of the globe (Mapillary, 2015).

This short video shows how it works:

Smartphones with an iOS, Android, or Windows operating system are not the only devices which support the capture of surroundings. Mapillary also offers tools which can be used to take shots with action camera’s such as Garmin Virb, GoPro and even drones, which is actually pretty cool (Mapillary, 2015).

Competitive advantages over Google Street View
Instead of solely focussing on digitally reconstructing the main highways and roads, Mapillary also tries to visualize those places that are inaccessible by car. Therefore, people who go on a hiking tour or an adventurous bike ride are encouraged to use the application, take photos and contribute to the community-driven photo representation of the world. By encouraging pedestrians and cyclists, Mapillary potentially can create an even richer view of the world and capture the places that remain untouched in Google Street View.

The other advantage of Mapillary over Google’s approach is that it can build upon more frequently updated content, resulting in a more real-time view. Depending on the place itself, it can take Google up to three to four years to update its Street View content. Mapillary on the other hand, has the potential to create a far more up-to-date representation and perform better at re-capturing the ever changing world around us.

The Rotterdam central station from a cyclist perspective in Mapillary

Active community
The community of Mapillary consists of passionate people who are working together to visualize the places they love. By doing so, they engage in a form of crowdsourcing. According to Tsekouras (2016), there are different reasons for people to engage in crowdsourcing initiatives, which can be classified into money, products, love and glory. In the case of Mapillary, money isn’t one of the reasons for community members to participate, since members won’t derive any monetary rewards from contributing. As far as I can see as an outsider, there is no list of top contributors in terms of meters mapped. Therefore, also glory will probably not be one of the main drivers for community members to engage in the act of crowdsourcing. Adding such a ‘top mappers list’ can have a positive impact on the amount of meters mapped by current members, and attract new people to become part of the Mapillary community in the future.

So, why do community members contribute? It’s relatively safe to conclude that members contribute in the interest of creating a new crowdsourced product, which is able to visualize the world around us. Besides, community members have a certain love for the places they go and want to share their experience with others. The following quotes from community members are good examples of the respective reasons to contribute to Mapillary:

“Compelling street view in exotic places around the world should be a crowd thing, a human thing, not a Google-only thing”

Kathleen (community member since August 2014, already mapped 1,695,999 meters)

“I like the idea that people can virtually visit the places I find beautiful”

Yann (community member since July 2015, already mapped 1,600,773 meters)

Shortcomings and future outlook
The potential drawback of crowdsourcing the act of taking photos that serve as input for a world-wide street view, is that the pictures will be taken with different devices. This results in difficulties in stitching together photos that differ in terms of image quality, not to mention variance in lighting, brightness and contrast. Jan Erik Solem, CEO of Mapillary, counters this argument by arguing that his company is ‘competing on reach and freshness, rather than quality, by going beyond major roads and cities’ (Solem, 2015). However, in this era of overly demanding users, I believe a relatively high quality standard for uploaded pictures is highly desirable.

It would be nice to contribute to the visualization of the ever changing world, by using the application during the upcoming BIM Eurotrip to Talinn in May and the BIM Study Trip to Silicon Valley in June/July this year. Hopefully, we can join forces, take on Google’s hegemony together and build a far richer competitor of Google Street View with at least some content generated by us as BIM students!

Robert Boer (356365)

References:

• Butcher, M. (2015) ‘Sequoia Invests In Mapillary To Crowd-Map The World Faster Than Street View’, January 15, 2015, URL: http://techcrunch.com/2015/01/15/sequoia-invests-in-mapillary-to-crowd-map-the-world-faster-than-street-view/ [Accessed: March 7, 2016]
• Mapillary (2015) URL: https://www.mapillary.com/how.html [Accessed: March 7, 2016]
• Mapillary (2015) ‘Meet community member Kathleen’, September 10, 2015, URL: http://blog.mapillary.com/members/2015/09/10/Meet-member-Kathleen.html [Accessed: March 7, 2016]
• Mapillary (2015) ‘Meet community member Yann’, September 10, 2015, URL: http://blog.mapillary.com/members/2015/09/10/Meet-member-Yann.html [Accessed: March 7, 2016]
• Lunden, I. (2016) ‘Mapillary raises $8M to take on Google’s Street View with crowdsourced photos’, March 3, 2016, URL: http://techcrunch.com/2016/03/03/mapillary-raises-8m-to-take-on-googles-street-view-with-crowsourced-photos/ [Accessed: March 7, 2016]
• Solem, J.E. (2015) Comment on: ‘Mapillary is Building a Crowdsourced Street View with User Submitted Photos’, URL: http://petapixel.com/2015/03/08/mapillary-is-building-a-crowdsourced-street-view-with-user-submitted-photos/ [Accessed: March 7, 2016]
• Tsekouras, D. (2016). ‘Customer Centric Digital Commerce 2016 – Lecture 3’, RSM Erasmus University