The first week of June was alive with Web 3 energy in Austin, Texas. Solana hosted Austin Hacker House
on June 4–9 and CoinDesk hosted Consensus
June 9–12. Nova Labs (formerly Helium) hosted Helium House ATX
Austin Hacker House | Panel: Building High Volume Data Applications with Solana
Hivemapper was one of the sponsors of Solana’s Austin Hacker House and Solana invited the Hivemapper engineering team to present a technical infrastructure overview. Some of our engineers presented “Building High Volume Data Applications with Solana,” hosted the Hivemapper Speed Hackathon
, and let their vocal cords loose at the Hivemapper Karaoke.
Hivemapper’s CTO & Co-Founder, Evan Moss, presented how the Hivemapper Network works
. He showed the power of Hivemapper’s decentralized approach with crypto incentives by showing a timelapse of 4K street-level imagery captured and uploaded at an intersection.
Hivemapper’s Head of Hardware Products, Ari Bruck, presented the tokenomics of the Hivemapper Network and its native cryptocurrency, HONEY. He explained how the Hivemapper Network computes map progress.
“The world is split up into regions and each period we compute how much map progress has been made in a region. Tokens are minted according to the HONEY token minting schedule and scaled according to map progress,” Bruck explained.
Map progress is computed as follows:
Coverage is the amount of unique surface coverage achieved
Activity is the volume of average activity throughput
Resiliency is a measure of how robust coverage is w.r.t. Activity
Ftr(ctr,atr,rtr) → [0, 1]
HONEY tokens are then divided among regions and distributed among map contributors based on weighting. Drivers with the Hivemapper Dashcam and Hivemapper Dashcam S are rewarded for the imagery they capture based on five factors:
Novelty is the inverse of saturation
Reputation is expected quality of content
Clarity is image quality
Consumption is royalty for consumer usage
Bounty is consumer injected demand
Collecting massive amounts of data to build a global map that truly provides freshness, quality, and coverage is a hardware, software, and network problem. “To address the hardware challenge, we’re building our own dashcams with multiple hardware partners in the United States and mitigating supply chain risk,” said Bruck.
The Hivemapper Dashcam and Hivemapper Dashcam S will collect the following core data:
GPS + Barometer
The engineers in the audience wanted to learn more about how Hivemapper’s software handles the massive amounts of data streaming into its network. One of Hivemapper’s software engineers, Jinglun Ding, presented how the Hivemapper Network processes massive amounts of data. The Hivemapper Network uses the in-house built Titan engine to carry out sensor fusion, ML/CV processing, crowd consensus review, continuity of network data across time and space verification, and ETL processes.
“The Titan engine processes sensor data and imagery collected with our dashcams and interacts with our on-chain program on Solana to reward contributors with our HONEY token,” Ding explained.
“We have created an idempotent account payment model for its on-chain payment program to ensure payment transactions are accurate and safe from any network failures, Ding said.
Austin Hacker House | Hivemapper Speed Hackathon
Hivemapper’s Speed Hackathon attracted many talented engineers who came up with interesting concepts and ideas of how to leverage the Hivemapper Network more broadly in the world. Jinglun Ding, Software Engineer at Hivemapper, judged all the submissions for the Hivemapper Speed Hackathon and selected the winning proposal.
Vera Wang and Michel Guan won the Hivemapper Speed Hackathon for their “Off-Set: Together, inspire 7 billion people to fight climate change” submission. In their presentation, they proposed integrating Off-Set into the Hivemapper Network to drive behavior that will reduce carbon emissions using crypto incentives.
Hivemapper Engineer Spotlight | Jinglun Ding
Jinglun Ding earned a Master of Software Engineering from Carnegie Mellon University and a Bachelor of Science in Computer Science, with minors in visual communication, and graphic design from Knox College. Before becoming a software engineer at Hivemapper, he worked as a Full Stack Software Developer at Crum & Forster in Richmond, Virginia, where he led and built a brand new insurance purchase and automated processing platform.
Jinglun distinguished himself at CMU by serving as team leader for his studio project — design and implementation of the mission control software to command and manage the MoonRanger robot, which is an 18kg mircorover to autonomously explore the lunar south pole for ice, volunteering to serve as the president of the MSELi, a student leadership initiatives, and working to create a supportive community during extraordinary times when many students were geographically dispersed.
Ding is the recipient of the 2021 James E. Tomayko Scholarship in Software Engineering. The award was established in 2006 to support students who demonstrate academic excellence, leadership potential, strong communication skills, and exceptional personal initiative. To date, 20 students in the Master of Software Engineering Professional Programs have received this scholarship.
Austin Hacker House | Hivemapper Karaoke
Solomona Phillips charmed everyone with her singing and won the Hivemapper Karaoke competition.
We’re hiring! Would you like to join our growing team? Take a look at career opportunities at Hivemapper