Maps are essential in our daily lives, whether we are navigating new cities or searching for the nearest coffee shop. Traditionally, maps were created by teams of cartographers and surveyors who would collect data and draw maps by hand. These physical maps eventually became the digital maps we know today.
Although maps have been digitized, OpenStreetMap and Waze still rely on thousands of humans to painstakingly edit and annotate the maps like you see below. At Hivemapper, we believe that there is a more efficient approach that involves combining AI and humans.
Caption: This GIF shows how an OpenStreetMap user edits the map. This tedious method will soon be replaced by AI.
Unlike Big Tech companies that take user data for free and don’t compensate map editors, the Hivemapper Network rewards contributors with HONEY tokens for collecting street-level imagery and refreshing it. Additionally, our collection device, the Hivemapper Dashcam
, is 1000 times cheaper than the Google Street View car.
Hivemapper’s current run rate of growth is roughly 8x faster than the 133,000 kilometers per month that Google Street View averaged in the first five years after the launch of Google Street View in May 2007. Hivemapper contributors have now collected 2 million km of unique street-level imagery
around the world, up from 1 million km
a month ago.
But, continuously updating imagery is just the raw ingredient of the Hivemapper map. The real magic is rooted in how we convert this imagery into an accurate and detailed map of our world – a $300B market
that is currently controlled by Google Maps.
Our Map AI processes the imagery drivers collect with the Hivemapper Dashcam and extracts map data: speed limits, turn restrictions, highway exit information, etc. With a constant supply of fresh imagery and this innovative Map AI process, the Hivemapper map is on the path to becoming the most accurate and detailed map of our world.
Help our Map AI get better at Object Detection and Classification
Map AI Trainer games help our Map AI produce quality map data. As contributors play these games, they train the Map AI on what a speed limit of 25 mph looks like or what a no u-turn sign looks like. Every object detection requires many contributors to review before reaching consensus on any given object.
Since releasing the Map AI Trainer game “Speed Limits USA” last week, contributors have completed over 130,000 reviews. Map AI Trainer reviews continue to grow exponentially.
Anyone can play AI Trainer Games and earn HONEY tokens
If you don’t have a car or can’t afford a Hivemapper Dashcam, you can still contribute to the Hivemapper Network and the future of map-making. The Map AI Trainer games are accessible from a computer or mobile device and are designed to be easy to learn and fun.
Map AI Trainer games are an innovative solution for building a map that is accurate, up-to-date, and accessible to everyone. By participating in these games, you can help improve the quality of our map data while earning HONEY tokens. Players are rewarded with HONEY tokens for quality submissions. If someone is just clicking and submitting nonsense reviews, that will quickly become apparent, and they will get blocked.
Start training our Map AI here.
Why Map AI Trainer Games are important for the Hivemapper Network
Hivemapper's rapid pace of collection has caught the attention of prospective customers, who are now interested in verifying the quality and precision of the extracted objects. In order to meet the demands of these customers, the Hivemapper Network must expand to accommodate millions of Map Editing and AI Training tasks every week, thereby ensuring the availability of an up-to-date and precise database of road objects.
For this reason, the Hivemapper Foundation has recently proposed to revise the weekly rewards formula to increase HONEY rewards for AI Training contributors by approximately 3.6x without impacting HONEY rewards for map contributors. Read more here
Future Map AI Trainer Games
Currently, the Map AI Trainer game “Speed Signs USA” is available. Soon, the Map AI Trainer games “Turn Restriction” and “Mount Placement” will be available. By playing these games, you can help improve the accuracy and quality of our map data while earning HONEY tokens.
Join the Hivemapper Community
If you have any questions about Hivemapper’s Map AI Trainer Games, join our Discord
and follow us on Twitter