Hivemapper Dashcam has started shipping
In preparation for the launch of the Hivemapper Mapping Network on November 3rd, the Hivemapper Dashcam
has started to ship to customers.
Hivemapper Dashcam worldwide testing
We’ve tested over 200 Hivemapper Dashcams across ten countries, including hot places like Las Vegas, NV. These dashcams have mapped 28,000 km so far with the longest drive at eight hours.
Pre-orders and shipping schedule
More than 5,500 dashcams, including the Hivemapper Dashcam and the Hivemapper Dashcam S, have been pre-ordered so far. We shipped out 300 Hivemapper Dashcams over the last week and will ship 3,500 more by the end of the year.
The mapping network’s three priorities
“We’re not the first company to build a dashcam. But when we designed the Hivemapper Dashcam, we asked ourselves how it will support the needs of the mapping network. Fundamentally, we’re building a dashcam to build a map. It has to prioritize quality from a mapping perspective,” said Ariel Seidman, CEO & Co-Founder of Hivemapper.
The Hivemapper Dashcam was built to support the mapping network’s three priorities: High-quality data, global coverage, and incredibly fresh data.
High-quality street level maps at low cost
On the right you see a street-level image collected by the Hivemapper Dashcams outside of Pittsburgh, PA. On the left you see a Google Street View image. These two images were captured at the same location. If you zoom in, you see the higher resolution of the Hivemapper image. It’s important to see signs to map, so we prioritized the lens resolution when designing the Hivemapper Dashcam.
The Hivemapper Dashcam generates accurate maps
The positional accuracy of all the data the Hivemapper Dashcam collects is less than three meters. It depends on where you are in the world. In certain locations, the positional accuracy is less than one meter. The other key metric we look at is durability. If you look at a 100 km drive, what percentage of that drive do you maintain that positional accuracy. Currently, we’re at more than 80%.
Where does our cost advantage come from?
A Google Street View car costs about $500,000 for the sensors, the car, the driver, insurance, and gas. The Hivemapper Dashcam costs $549, which means that for the price of one Google Street View car you could have approximately 1,000,000 Hivemapper-equipped cars.
With far more vehicles collecting, this means that Hivemapper can see a location about 24 times more frequently than Google Street View. Additionally, this means that Google Maps only gets 1 shot on goal while Hivemapper gets 24-48 shots on goal about every two years.
“If you look at hockey, the team that gets more shots on goal usually wins. If a Google Street View car passes a low lit road, they might not pass that location for another two years. Whereas if Hivemapper doesn’t get good lighting one day, it will see it another time,” said Seidman.
Case Study: Manila, Philippines
Manila, Philippines has about 11,613 road km. In its alpha stage, the Hivemapper Network covered 95% of all the roads in Manila and mapped 180,000 road km while it was active. 75% of the Manila metro area was getting refreshed every month.
Case Study: Los Angeles, USA
At 94,256 road km, the Los Angeles metro area is more than 8 times bigger than Manila. Hivemapper reached 56% coverage over 6 months and was refreshing 30% every month. During this Hivemapper Dashcam alpha stage, we were using a third-party dashcam. Contributors had to take out the SD card every three days and manually upload it through their computers. The Hivemaper Dashcam will automatically upload imagery to the Hivemapper Network, making it a passive mapping experience.
“We’ll present more data like this in the coming weeks. The one big learning from this is that once we got hooked into the Uber and Lyft drivers and other professional drivers, then the coverage took off,” said Seidman.
The Hivemapper Dashcam is designed for privacy and global mapping
Viewing the Hivemapper Network as a two-sided marketplace, we designed it so it would meet the needs of both contributors (supply) and customers (demand). We optimized the Hivmeapper Dashcam for data consistency, passive contributor experience, build for vehicles, works in extreme conditions, and data everyone trusts.
“There have been other projects that have used iOS and Android devices and the big problem with that is that there are a lot of different Android devices and there are a lot of different iOS devices as well. And they all have different camera specs. Because they all have very different camera specs, they produce a very wide range of imagery. There needs to be an ML processing engine that’s consuming the imagery or a map editor or analyst who’s consuming the imagery. Dealing with this variety is tough. Whereas with the Hivemapper Dashcam and other dashcams to follow there has to be, and there will be, network specifications that will ensure data consistency,” said Seidman.
Passive experience for contributors
The Hivemapper Dashcam comes with a mount and power cord. This video
shows how to install a Hivemapper Dashcam. Once the Hivemapper Dashcam and the Hivemapper App are paired, all imagery will be uploaded to the network automatically via WiFi.
Designed for privacy
“Today, we mask faces and license plates. We do that on the server side. Ultimately, we want to do that on the dashcam level as well so that we never have to deal with any private information,” said Seidman.
A dashcam device we all trust
“If you’re consuming all this imagery and map data, you have to be confident that it was collected on 2nd and Mission St and not rural Kansas. We’re not naive about the fact that people will try to spoof the network and they will try to pretend like they’re driving on 2nd and Mission St. So there are a few mechanisms we have in place like location verification using the Helium network, encrypted GPS using u-blox, and Map Image QA,” said Seidman.
The Hivemapper Network launches on November 3, 2022
“The Hivemapper mapping network goes live on November 3rd along with HONEY tokens. We’ll open Map Image APIs to customers we’re working directly with, and we’ll fulfill dashcam pre-orders,” said Seidman.
Where are we headed?
“The dashcam is definitely the primary pillar of the mapping network. It collects imagery and location data. Over time, we will build other types of sensors to collect different kinds of data for other layers of the map. One is a 360 camera and another is a traffic puck that pushes GPS coordinates to build out a traffic layer,” said Seidman.