How the Hivemapper
Network works

Updated on March 28, 2022
24 minute read


Mapping Economy
Billions of people around the world depend on maps each day. Maps are used by insurance providers, real estate services, logistics companies, navigation and delivery apps, and governmental organizations, just to name a few. Maps are an essential part of the world’s technology infrastructure and represent a $300B market.1
The Problem
Today, global maps are largely controlled by a few companies because they are extremely expensive to build. This has led to numerous problems with today’s maps:
  • High costs for businesses: The cost of mapping APIs continues to increase due to the lack of competition.2
  • Uneven freshness and coverage: These maps are not as fresh as they should be due to the incredible expense required to build and maintain them. Developing markets have vastly inferior maps given that the high cost to map breaks the economic models.
  • Free use of user data: Existing maps use our private and sensitive location data to improve their own maps without compensation.
  • Prone to censorship: Big tech companies operate a wide set of businesses in many countries, and are often forced to censor or modify their maps to satisfy politicians and land owners.
The Benefits of Building a Mapping Network on the Blockchain
Blockchain and crypto incentives solve these problems by dramatically reducing the cost to map on a global scale, while rewarding contributions to the map and promoting freshness and uncensored quality.
High quality 4K dashcams are widely available for only hundreds of dollars and can be connected to software for efficient collection of 4K street-level imagery. Map QA reviewers, annotators, and annotation QA reviewers are incentivized with ownership to maintain the quality of the map. Additionally, thanks to the decreasing cost of machine vision compute cycles it is now feasible to transform imagery to valuable map data on a global scale.
Imagine a network of map contributors and map consumers intricately connected on a blockchain, participating in the exchange of valuable map data. Map contributors build and refresh the map by capturing 4K street-level imagery, carrying out quality assurance, and annotating imagery. Map consumers leverage the map via a set of APIs such as images, direction, geocoder, and more.
With the introduction of a decentralized mapping network and cryptocurrency, the mapping network injects decentralization and built-in crypto incentives into an industry currently controlled by monopolies and governments that take our data for free.
Hivemapper’s decentralized global map and cryptocurrency
The Hivemapper Network is a system that incentivizes map coverage, freshness, and quality with ownership. By installing a simple 4K dashcam on a car or truck, a contributor can earn a new cryptocurrency, own part of the decentralized global map, and support the world’s critical geospatial infrastructure in a cost-effective way.
Powering the Hivemapper Network is the decentralized global map on the blockchain and its cryptocurrency HONEY. With the introduction of a decentralized mapping network and cryptocurrency, the mapping network injects decentralization and built-in crypto incentives into an industry currently controlled by monopolies and governments that take our data for free. The result is a state-of-the-art map of our world that is constantly-renewing, high quality, truly covers our whole world, and is collectively owned by its contributors.
Token Economy
Fig. 1 / The Hivemapper Network

This diagram shows the two-sided marketplace between map contributors and map consumers interacting on the Hivemapper Network with its cryptocurrency token HONEY.

In the following sections, we describe how the Hivemapper Network is designed. We will cover the milestones necessary to launch and scale it, the tokenomics of HONEY, the burn and mint system with a net emissions model, and how the system will incentivize map coverage, freshness, and quality.

System Overview

Map Tiles
The global map is divided into small hex shaped tiles referred to as map tiles. The map tiles are the atomic unit of the map and based on H3 cells. We intentionally use small H3 cells as the basis for our map tiles (shown in the graphic below) to make it easy to start contributing and to avoid issues with “partial” tiles contributed.
Tile Sizes
Fig. 2 / Map Tiles

Map tiles are the atomic unit of the Hivemapper map that contributors use to build and refresh coverage. Each map tile is hex shaped. Trillions of these hex shaped tiles cover the entire earth.

4K Street-Level Imagery to Map Tiles
The 4K street-level imagery and related GPS metadata collected with approved dashcams and the Hivemapper app serve as the raw ingredients for the global map. This data provides coverage for map tiles as seen in Figure 3 below.
Tiles Mapped
Fig. 3 / Map Tiles Along a Path

Map tiles that have been covered on the road are represented in pink, as shown by H3 cells.

Map Contributors
The collective work of map contributors on the Hivemapper Network results in a high quality global map. The chart below details the different types of map contributors and the type of work they do on the mapping network.
ContributorWork They Do
DriversCollect 4K street-level imagery via a supported 4K dashcam and Hivemapper app
Imagery QA ReviewersValidate the 4K street-level imagery collected by drivers
AnnotatorsAnnotate the map with details such as street direction and name. Annotations can be broadly defined; for example, a customer may pay to add artwork associated to the map
Annotation QA ReviewersValidate the work of Map Annotators
Software DevelopersExtend the protocol, product, and work on core mapping features and the Hivemapper app
The Hivemapper Dashcam
In the summer of 2022, Hivemapper will begin selling the world's first crypto-enabled dashcam. The Hivemapper Dashcam is an open source dashcam that natively integrates with the Hivemapper Mapping Network through a seamless mobile app.
The Hivemapper Dashcam is based on the Open Dashcam specifications that ensures:
  • Location Authentication: Multiple layers of security to ensure that the dashcam is authentically geolocating its position
  • Automatic data transfers: Automatically transfers the collected data from the dashcam to the Hivemapper Network via integration with the Hivemapper Contributor App for iPhone and Android
  • Dynamic data collection: Dynamically determines the data required for the map - ignores the rest
These capabilities make it the ideal dashcam for mapping on a decentralized network.
Hardware manufacturers can incorporate the Open Dashcam specifications in their own dashcams, and seek approval from the Hivemapper Foundation to run their compliant dashcams on the mapping network.3
Fig. 4 / Hivemapper Dashcam

The Hivemapper Dashcam is an open source camera based on the Open Dashcam specifications that is optimized for collecting imagery from a vehicle for the purpose of mapping.

HONEY is the Hivemapper cryptocurrency of the decentralized mapping network. The Hivemapper Network is built on top of the Solana blockchain.
Map contributors mine HONEY by contributing to the Hivemapper Network. A fixed number of HONEY tokens are minted decreasingly over time. The maximum number of HONEY tokens that will ever exist is 10 billion.
Map API Services
APIs built on top of the global map enable developers to cost effectively integrate maps and geolocation services into their applications. Today, Hivemapper Inc. offers individuals and organizations the powerful Map Image API. The set of APIs that Hivemapper Inc. and other organizations can build and commercialize on top of the global map includes Driving Directions, Geo Search, Traffic, and more.
The Hivemapper Foundation ensures that any company or organization can build and commercialize API services on top of the global map.
Map Consumers
Customers who want to integrate Hivemapper’s map APIs into their applications can purchase API calls with Map Credits. Map Credits are created by burning HONEY tokens, ahead of any use of map APIs. These burned tokens then increase the number of tokens available to mint and to pay to map contributors.
The Burn and Mint Equilibrium
The Burn and Mint Equilibrium with Net Emissions model is used, such that whenever map consumers burn tokens to access the network, an equivalent number of tokens is added back into the rewards pool for map contributors as shown in Figure 5 below.
Hivemapper Customer Flow
Fig. 5 / The Burn and Mint Equilibrium

When map consumers use map APIs this transaction burns tokens, and an equivalent number of tokens is added back into the pool to reward contributors.

At a high level, contributors, including drivers, annotators, and QA reviewers, build the database of fresh and accurate map data in exchange for HONEY. Consumers of the data pay into the network by purchasing and burning tokens to compensate the contributors for their work.

Upcoming Milestones

Hivemapper Network: Testnet (Alpha)Participants will be paid in cash, and will also receive a free Hivemapper Dashcam4Summer 2022
Hivemapper Mapping Network FoundationFormally constitute the Mapping Network FoundationSummer 2022
Hivemapper Dashcam begins shippingInitial orders of Hivemapper Dashcam shipSummer 2022
Hivemapper Network:
Mainnet (Beta)
  • Expand to ~30 regions
  • HONEY token becomes only form of rewards5
2H 2022
Hivemapper Network:
Global Launch
Hivemapper HONEY tokens begin expanding globally2H 2022
Region Rollout
In the beta phase, contributors can map and QA anywhere, but they can only earn rewards in certain regions. Regions are defined as broad areas that go well beyond a city center. For example, a small set of current regions can be viewed here. To ensure map quality at scale, the Hivemapper Network will activate new regions in a phased approach as outlined in the schedule below. Given the scale of the mapping network, dates below may be adjusted.
HivemapperHivemapper Network
PhaseCurrentTestnet (Alpha)Mainnet (Beta)Mainnet (Global)
TimeframeActiveSummer 20222H 20222H 2022
Open Metro Regions
  • Dallas, USA
  • Kuala Lumpur, Malaysia
  • London, UK
  • Los Angeles, USA
  • Manila, Philippines
  • New Orleans, USA
  • Palawan, Philippines
  • Seattle, USA
  • Singapore
  • Dallas, USA
  • Kuala Lumpur, Malaysia
  • London, UK
  • Los Angeles, USA
  • Manila, Philippines
  • New Orleans, USA
  • Palawan, Philippines
  • Seattle, USA
  • Singapore
  • Austin, USA
  • Bangkok, Thailand
  • Barcelona, Spain
  • Boston, USA
  • Chicago, USA
  • Dallas, USA
  • Istanbul, Turkey
  • Jakarta, Indonesia
  • Kuala Lumpur, Malaysia
  • Lagos, Nigeria
  • Lisbon, Portugal
  • London, UK
  • Los Angeles, USA
  • Madrid, Spain
  • Manchester, UK
  • Manila, Philippines
  • Mexico City, Mexico
  • Miami, USA
  • Nairobi, Kenya
  • New York City, USA
  • View More
Anywhere in the world
Fig. 6 / Region Rollout for Hivemapper Decentralized Mapping Network and HONEY

Contributors in the testnet phase will continue to receive cash payments; once the Hivemapper Network launches on the blockchain rewards will only be made in HONEY tokens.

Mapping Network Foundation

In 2022, the Hivemapper Mapping Network Foundation will be formed, and over time will take on many of the responsibilities currently held by Hivemapper Inc.6 The Mapping Network Foundation is a global not-for-profit organization that governs and maintains the Hivemapper decentralized mapping network and the underlying open mapping technologies.
As core digital infrastructure that billions of people rely upon, the map should become a public good owned by its contributors and governed by a foundation with a clear mandate
Why a foundation? As core digital infrastructure that billions of people rely upon, the map should become a public good owned by its contributors and governed by a foundation with a clear mandate to proliferate a fresh and global map accessible to all. By transferring many of the key governance responsibilities and underlying mapping technologies for building and annotating the map to the foundation, and by making the technologies related to the dashcam and data processing open source, the mapping network can serve far more people across the globe, and do so for far longer.
MissionThe foundation is a global not-for-profit dedicated to the proliferation of a fresh and global map as a public good
StructureThe foundation will be established in 2022 as a US based entity with three initial committees: technical, map quality, and economic, and will host monthly calls with the community
Responsibilities(1) Ongoing Map Collection, Quality, and Annotation Activities. (2) Establishes a transparent process for making updates to the underlying protocol and formulas that are described in this document. (3) Manages and maintains the open technologies that support the Hivemapper Mapping Network
PartnershipsThe foundation will work with hardware manufacturers to create compatible Open Dashcam devices and establish a certification process for adding new hardware devices to the network


The HONEY Token
HONEY is the cryptocurrency of the Hivemapper blockchain built on top of Solana, which rewards contributions to the mapping network. Mined HONEY tokens will be available in a Hivemapper Web3 wallet. From there, the user is free to transfer their tokens on the Solana blockchain wherever they please; inclusive of other wallets and exchanges. There is a fixed supply of 10 billion HONEY tokens minted on a set schedule described in the Token Economics section.
The fundamental components of the ecosystem are map coverage, QA, and annotations. This is where the bulk of the token rewards will be distributed. The Hivemapper Network Foundation will use the reward structure described below to determine the number of tokens that contributors are awarded each time tokens are minted, including a large and increasing portion for Map Consumption Rewards.
TypeDescriptionWho earns this reward?Atomic UnitInitial % of allocation
Map Coverage & QualityThe most basic kind of reward for collecting quality imagery that successfully covers a map tile. Quality checks also draw from this bucket as needed.Drivers and Map Quality ContributorsMap Tile70%
Map ConsumptionWhen a map tile is consumed via any API, it earns incremental rewards.Drivers and Map EditorsMap Tile20%
Map Annotation & QualityProvides annotations for a map tile. Annotation quality checks also draw from this bucket as needed.Map Annotators and Map Quality ContributorsFeatures0%
Map ValidatorsValidate the quality signals, must stake to become a validator.Crypto Network ContributorsTBD5%
Map ProcessorsProvide resources for operating the network infrastructure.Crypto Network ContributorsMap Tile5%
HONEY Token Distribution Over Time
Initially, collecting imagery to build Map Coverage is critical. Over time, the focus shifts to Map Annotation and Map Consumption as the map matures. The ‘clock’ is driven by the token minting schedule defined below which is controlled by overall map progress, an indicator of coverage and maturity.

HONEY Token Distribution over Time

Line chartSpreadsheet
Fig. 7 / Token Distribution Over Time

The allocation of tokens for different types of work on the mapping network shifts as the map progresses and matures. The speed at which it shifts will depend on the rate of map progress across the network.

Networks Dynamics
The Hivemapper Network will incentivize and manage various behaviors on the network. These all relate to ensuring the health of the mapping network and quality of the map.
The system of incentives that is set up is critical to creating a healthy and valuable network for both customers and contributors. There are three primary dimensions along which customers value maps, and therefore along which the HONEY token will pass incentives on to contributors: Coverage, Freshness, and Quality. The sections below describe the specific incentives that the Hivemapper Network will use to drive map Coverage, Freshness, and Quality.
There are three primary dimensions along which customers value maps, and therefore along which the HONEY token will pass incentives on to contributors: Coverage, Freshness, and Quality.
Specific weightings, numbers, and functions detailed in the section above and below will be finalized and adjusted by the foundation in an open and transparent process based on balancing the economics of the system.

Map Coverage

Coverage Freshness Quality Chart
Fig. 8 / Map Coverage

Map Coverage is incentivized by rewarding route novelty, region demand, map consumption, and other signals.

It’s important to generate broad coverage, while tailoring coverage to areas where customers want to use maps. Generating broad coverage builds a base which then attracts more customers; tailoring to customer needs brings in ongoing funding (via customer purchases) to support contributors as they continue contributing. To those ends, we will use the following incentives to drive coverage:
Route Novelty
Route novelty promotes a focus on less densely mapped areas, rebalancing rewards towards under-mapped locations in order to drive broad map coverage. As the network grows initially, this will be one of the strongest factors in driver token earnings.
Initially, token rewards for all tiles in a region are equally distributed. As certain routes become saturated, the Novelty Multiplier and therefore the rewards for tiles across those routes will decrease. Meanwhile, rewards for tiles across routes that are under-mapped will increase, incentivizing drivers to alter their routes accordingly.
For example, if Van Ness Avenue starts to get over-mapped, the protocol will lower the token rewards for map tiles on that road. Map tiles on under-mapped roads will be relatively more desirable, leading drivers to adjust their routes when possible. Route Novelty is calculated separately for front-mounted and side-mounted dashcams.
Route Novelty
Fig. 9 / Route Novelty

Route novelty rebalances rewards towards under-mapped locations in order to drive broad map coverage.

Region Multiplier
Region multipliers allow anybody to focus mapping in areas where there are current or expected use cases for map data. As the map is initially developed, this will have a strong impact on driver token earnings.
The default region multiplier is 0.1. Once the blockchain is live in a region, the protocol does not support turning off HONEY rewards to mitigate the risk of censorship. However, the region multiplier can be increased or decreased to reflect focus in an area.
The protocol will have the ability to set higher rewards up to a multiplier of 1 for specific regions based on consumer demand and economic activity. For example, the protocol can adjust the region multiplier upwards for a region based on high GDP growth, population growth, building permits, and other publicly available economic statistics that are proxies for physical changes and economic activity within a region. Maps are most critical in areas where high amounts of growth and change are occurring — the protocol will prioritize accordingly.
A map consumer can additionally burn tokens to increase the total rewards distributed for a region. The number of additional tokens distributed for that region is equal to the number of tokens burned.
Team Rewards
Achieving map coverage goals can be incentivized via Team Area Completion bonuses, which are shared by all who assisted in 100% completion of an area.
These rewards are funded by a customer – e.g. the customer pays to create a bonus for completing all of Manila, and every driver who contributed splits that reward if achieved.
Map Consumption Rewards
Contributors who map tiles that are then used by a customer will earn the additional Map Consumption reward. While this is initially a smaller portion of driver token earnings, over time the value of this incentive and the focus on it will increase; in the long term, this will be a strong factor in driver token earnings.
Based on the time series of map additions, at the time that a given route is utilized by a customer, the associated rewards are split amongst all who contributed that route. The highest reward is given to the most recent contributor, with proportionally lower scaled rewards for less recent contributors.
  • For each Tile, Map Consumption Rewards are distributed with the greatest rewards going to the most recent contributor in that location, and extending back at fractions to the 10 most recent contributions.
  • The same contributor can be included in the list multiple times.
  • There is no cutoff for freshness.
See below for an example in practice.
While the below example governs most use cases, there are times when a customer is interested in very specific data, such as how a road looked on a certain date. If the customer pays for this very specific data, then the Map Consumption reward will be paid entirely to the contributor who collected this data.
Location: City Hall, Polk Street, San Francisco – Recent mappers as of 6/1/2022 0:00
nthMost recent mappersDate/timeReward portionNotes
77joyful_purple_chicken5/31 5:12pm0.5
76dancing_zinc_tardigrade5/31 11:04am0.25
75calm_sky_reindeer5/30 12:15pm0.125
74joyful_purple_chicken5/28 4:46pm0.0625
73narrow_raisin_chipmunk5/26 1:44pm0.03125
72narrow_raisin_chipmunk5/23 12:42pm0.015625
71jovial_amber_raccoon5/21 1:19pm0.0078125
70joyful_purple_chicken5/20 3:30pm0.00390625
69dandy_green_rat5/19 10:19am0.00195312
68breezy_mango_boa5/16 8:10am0.0009765610th most recent
67fierce_blue_dinosaur5/15 3:42pm
66narrow_raisin_chipmunk5/14 3:05pm
Annotation Rewards
Map annotation (or map editing) adds another layer of coverage to the map. Map editors who annotate the map will earn HONEY tokens for contributions.
Map annotation is a future area of focus. Editing traffic signs, building boundaries, and street names are examples of map edits.
Exact schedule for annotation contributions will be developed; currently they do not receive rewards.
  • Rewards for Map Editing should happen at the granular level e.g.
    • Add/Edit 1 traffic light = 0.x HONEY token
    • Add/Edit 1 road segment = 0.0x HONEY token
    • And so on
  • These rewards are granular to provide control over specific features (e.g. want more roads and less traffic signs). Makes it clear what you need to edit
  • Quality will be tracked with a quality score, based on the signal of how often people’s work is challenged by other editors. Future edits and reversion of work do not retroactively affect tokens earned, but they do affect quality scores and reviews going forward.
  • New map editors will not be allowed to make large numbers of map edits until their quality score is trusted.
  • The rewards for Map Editing are lower than driving with a dashcam. Driving is more expensive and requires expensive equipment — a vehicle and dashcam.
Driver Staking Fees
To ensure ongoing coverage, the network should ensure that existing contributors are rewarded enough to maintain their interest. Hence, the network can limit oversaturation of a given region by using driver staking fees.
The initial staking fee will be set to a specific amount of HONEY tokens, which can be adjusted through governance as network dynamics are observed over a longer period of time. Hivemapper, Inc. and other entities can elect to pay contributors’ staking fees in certain regions for limited periods of time, to promote increased map coverage in specific regions.
In addition to the effect on network coverage, driver staking fees tie drivers' interests more strongly to the success of the network. This means that there is a deterrence to fraudulent activity and an incentive to ensure that value is maintained through high quality data.

Map Freshness

Freshness Multiplier Chart
Fig. 10 / Map Freshness

Map Freshness is incentivized by increasing the value of tiles that need to be refreshed with the freshness multiplier.

Customers want data that is fresh enough to be reflective of the current on-the-ground reality. However, there is also a limit to how frequently refreshed data is useful. Therefore, the network will incentivize updating data on a preferred cadence.
Freshness Multiplier
The goal of the freshness multiplier is to reflect whether a tile is getting stale (and therefore more valuable to map) or already fresh (and not currently valuable to map).
One challenge is knowing when a tile should be considered “stale”. Many customer use cases would be satisfied with data that has been refreshed within the past week, so by default the mechanism will target weekly data refreshes. However, there are also several potential ways to incentivize faster refreshes as well:
  • Customers can spend tokens to increase the total rewards for collecting an area to match their use cases.
  • Public transportation organizations can be granted the ability to temporarily flag areas for a quick refresh to address urgent disaster or safety needs.
  • Mechanisms may be set up to allow contributors to report issues that require refresh e.g. closed roads. This will be paired with a reputational/token-tied consequence for any abuse of the reporting system.
Mechanically, this mechanism is a multiplier between 0 and 1 that scales per-tile payouts. For an area that has been recently mapped, its value drops to 0, and grows back to 1 by the next time limit. The contributor earns the Multiplier that existed at the time that they submitted their map. Each time that the area is mapped, the clock resets. Freshness is calculated separately for front-mounted and side-mounted cameras. See the graphic below for an example of how the multiplier changes as time passes and as tiles are mapped and re-mapped.
Fig. 11 / Map Freshness

The Freshness Multiplier is set between 0 and 1, and is used to increase the value of a tile that needs to be refreshed. The freshness multiplier incentivizes map freshness.

Map Quality

Map Quality Chart
Fig. 12 / Map Quality

Map Quality is incentivized by rewarding data usefulness, clarity of view, fewer QA reviews, and higher contributor reputation score.

For data to be maximally useful, it must be of reliably high quality.
There are broadly three potential map quality issues:
  • Issues in setup – Data collection devices are not properly installed. For example, a new contributor whose dashcam is mounted incorrectly, leading to obstructed view or solar glare.
  • Correct setup, but unusable data – Uploaded data that can’t be used due to the following issues: Nighttime, excessive rain, excessive other vehicles blocking view of the road, etc.
  • Intentional spam – People uploading or annotating data that is duplicate, fake, altered, etc. with the hope to add it to the map and/or earn token rewards.
There are two primary types of feedback signals the network uses to detect these issues:
  • Human feedback – Provided by humans working on QA tasks on the network.
  • Machine feedback – A wide variety of detection can be used to generate quality signals, as well as to auto-screen data that is clearly not valuable (e.g. all-black nighttime data). Examples of such feedback include location attestation using GNSS, LoRa (location triangulation using Helium blockchain), and other on-board sensor data.
Considering potential quality issues and methods of detection, the network will implement the following mechanics:
Data Usefulness
Data Usefulness
Map Quality Signals
Human + Machine Feedback
  • Authentic
  • Daytime
  • Occlusionetc.
Distribute Token
Fig. 13 / Data Usefulness

Data Usefulness is assessed via human and machine feedback followed by the Validators to further verify quality signals. Validation happens before rewards distribution and contributor reputation score updates.

Data deemed unusable for purposes of mapping (whether via human or machine feedback) will not earn rewards. For example, if the dashcam’s view is heavily occluded by other vehicles on the road or if the imagery was collected at night, this data will be excluded prior to distributing token rewards.
The Hivemapper Foundation map quality committee will review and approve new map quality signals or tune existing signals. The Validators will then verify signals that are sent to it; validation happens before rewards distribution and contributor reputation score updates.
The details of the quality signals used will be governed by the foundation. Broad categories of map quality signals include:
  • Mount position detection - is the camera mounted properly enabling quality imagery to be collected
  • Integrity – detecting issues with fidelity, glare, lighting, or obstructions
  • Authenticity - is the imagery authentic to its asserted location and is it consistent with imagery uploads in the same area
Quality signals can be defined to apply per-region or per-time period.
Clarity of View
Clarity of view

Front-facing, interior


Front-facing, exterior

Fig. 14 / Clarity of View

Rewards superior clarity of view and therefore higher quality 4K street-level imagery for the map.

Imagery that provides a clearer view earns a higher multiplier than imagery that does not.
This will typically mean that an externally mounted dashcam would earn more token rewards than a dashcam mounted inside a vehicle which could be subject to glare from a windshield unless carefully positioned.
This incentive can also be used to support future areas of interest for the mapping network including support for externally mounted 360 degree cameras.
QA Tax
QA Tax Graphic
Fig. 15 / Review Tax

Contributors with high reputation scores retain the vast majority of their rewards, as their work is not subject to significant amounts of QA; contributors with low reputation scores must pay a significant percentage of their token rewards as a QA tax.

It’s essential that there is a robust feedback loop to support a high map quality. To ensure sufficient QA volume, QA work will be funded via a tax on contributors whose contributions require QA reviews.
When a map contributor is new, a QA Review Tax will be assessed on a high percentage of their work. As they prove their reputation over time, the percentage drops systematically, but will also spike much higher if issues are observed.
The system is balanced, because the QA Review Tax funds the QA work, and the QA work directly benefits the contributor who is being taxed. It gives them valuable feedback if there are issues, and unlocks faster rewards when their quality score improves.
It is important to ensure that QA reviewers themselves are doing good work. Therefore, QA reviewers also have a trust score. When they are new or have a lower trust score, all of their work is also sent to other QA reviewers. QA reviewer quality will be based on their % agreement in these checks, and will be used to determine the need for further QA checks. When QA work is sent to multiple people, they split the rewards for that review.
QA Rewards
The QA reviewers receive rewards based on QA’ing contributions that drivers submit. The work available to the QA’ers is controlled by a queue based on drivers’ amount of work submitted and drivers’ current reputation scores.
QA reviewers also have a QA score. QA tasks will periodically be sent to multiple QA reviewers, and the rate of agreement will be used to assess how reliable each QA reviewer is. A QA reviewer with low agreement will not be able to individually review work. QA reviewers who cannot individually review work will receive a lower token reward for completed tasks, due to not being the sole reviewer of the task.
Annotation QA Rewards
QA for Annotation tasks is parallel to QA for driver mapping contributions.
The QA reviewers receive rewards based on QA’ing annotation tasks. The work available to the QA’ers is controlled by a queue based on annotators’ amount of work submitted and annotators’ current reputation scores.
QA reviewers also have a QA score. QA tasks will periodically be sent to multiple QA reviewers, and the rate of agreement will be used to assess how reliable each QA reviewer is. A QA reviewer with low agreement will not be able to individually review work. QA reviewers who cannot individually review work will receive a lower token reward for completed tasks, due to not being the sole reviewer of the task.
Contributor Reputation Score
Each contributor has a contributor score that will start out neutral, and improve or fall. A high reputation score leads to benefits, such as faster turnaround on token earnings, higher total earnings, and unlocking portions of tokens to sell.
When calculating the contributor reputation score, we take into account the quality signals described above (particularly those around setup issues or fraud), as well as the reliability of the dashcam that the contributor uses.
One key principle is that it’s hard to build a reputation, and reputations are fragile; bad work more quickly affects a contributor’s reputation than good work does. So, negative feedback has a bigger impact on the contributors’ quality score than positive feedback. In cases of clear fraud attempts, there will be severe and immediate consequences to a contributor’s reputation and quality score.
An existing contributor who begins mapping in a new region will begin with an average of the standard ‘new contributor’ score and their earned score from another region.
Tiles Mapped
Fig. 16 / Contributor Reputation Score

The Contributor Reputation Score is calculated using the quality signals described in the 'Map Quality' section.

QA Score Token Holdback
Some rewarded tokens will be unable to be accessed pending achieving a satisfactory QA rating.
A token holdback occurs when a contributor has theoretically earned a certain number of tokens, but the Hivemapper Network escrows some or all of the tokens pending certain qualifications that must first be met. The exact blockchain mechanism for token holdback and release is yet to be defined.
One example of a token holdback is a QA score holdback. Contributors who have not yet demonstrated an acceptable quality score will be subject to token holdback. The required score to clear the holdback is to be determined, but will be higher than the neutral starting score that a contributor receives when joining.

Token Economics

Fixed Token Supply
There will be a fixed maximum supply of 10 billion HONEY Tokens. To ensure this fixed supply, a fixed number of HONEY tokens are minted decreasingly over time. The minting schedule is described in the section below.
Of the 10 billion tokens, 20% will be allocated to investors who provide startup capital, 20% to employees of Hivemapper, Inc. who build the technical and operational systems required, 15% to Hivemapper, Inc. for providing ongoing security and technology, and 5% to the foundation that will be established to govern and facilitate the ongoing success of the network. The largest portion, 40%, will be allocated to token rewards for ongoing contributors to the map ecosystem as described in this document.
Token Minting Schedule
When the mainnet launches, the Hivemapper Network will begin minting and distributing 4 billion HONEY tokens to contributors. The minting rate will decrease by at most 40% every two years, as shown in the schedule below. This schedule is the maximum rate at which token minting can occur; however, minting will almost certainly occur at a slower rate, as it is driven by map progress.

HONEY token minting schedule

Download CSV
HONEY Token Minting Schdule
Fig. 17 / Maximum HONEY Token Minting Schedule

When the map is progressing at maximum speed, the minting rate will decrease by 40% every two years. A total of 4 billion HONEY tokens will be minted.

When the map is progressing at maximum speed, the minting rate above is achieved. However, when map progress slows, the overall minting rate will slow, and the date when the minting rate decreases will be stretched out. We do not expect the minting rate to exactly match the schedule above. This dynamic approach is designed to ensure that early contributors are rewarded generously, but rewards are never overly concentrated in regions that are not adequately producing map coverage.
For the minting to progress at maximum speed, all regions that make up the global map must meet a minimum threshold of map coverage, activity, and resiliency each week. Each region’s contribution to the overall minting progress is weighted by a factor representing its importance to the global map. Map Progress7 in a given region is determined by three progress factors: Coverage8, Activity, and Resiliency, defined below.
Region Map Progress
Fig. 18 / Map Progress Factors

The minimum thresholds that a region must meet to mint the maximum potential tokens in a given week are based on Coverage, Activity, and Resiliency.

Historical data from the mapping network (to be published later) strongly suggests that regions meetings these minimum progress factors are on a path towards building a complete, fresh, and high quality map resilient to contributor churn.
When a region meets the minimum threshold for the past week it will earn the maximum potential tokens for the week, and the HONEY token rewards will be distributed to contributors in that region based on their contributions. If a region does not meet the minimum threshold, then HONEY token rewards are distributed based upon the Map Progress and Region Multiplier formulas. Our historical data suggests:
  • Large and sprawlings regions like Los Angeles take longer to achieve minimum map progress thresholds
  • Compact and dense regions like Manila achieve minimum map progress thresholds relatively quickly
  • Once a region achieves the minimum map progress threshold, it generally continues to do so in future weeks
How Map Progress Works
Fig. 19 / How Map Progress Works

Two out of the four regions did not meet the minimum map progress thresholds and instead of earning the maximum of 25 tokens earned 10 tokens each.

For the purposes of explaining this concept, assume for a moment there are only four equally weighted regions in the world as seen in the graphic above, and the maximum weekly minted tokens is 100. When all regions meet their minimum map progress thresholds in a week, 25 tokens are distributed to each region and the map is progressing at its maximum speed.
Now consider a case where two out of the four regions achieved the minimum map progress thresholds described above. The token distribution for those regions not meeting their minimum thresholds is determined by the Region Multiplier and Map Progress formulas. In this example, they each earned 10 tokens.
Only 70 of the 100 tokens were distributed that week, and the treasury retains 30 tokens. These tokens will ultimately be distributed as the map progresses faster.
The choice to anchor token minting rate to map progress within a region not only ensures that tokens are disproportionally distributed to higher quality regions, but it also encourages contributors to recruit new contributors within their region to ensure map progress thresholds are met for their region.
Burn and Mint System with a Net Emissions Model
The Burn and Mint structure creates an on-chain marketplace between product creators and consumers. Consumers buy and burn the token in order to get access to goods and services – in Hivemapper’s case, collected data, or the ability to set priorities for upcoming data collection. In particular, the mechanics are such that they buy and burn existing tokens, giving them credits. When a customer uses their system credit, their usage is tracked so that specific creators can be allocated associated rewards.
Given that tokens will be burned over time, and we do not want to deplete the supply of tokens in the system, we will use a Capped Net Emissions model, as modeled previously by Helium in HIP 20.9 This means that the net number of tokens burned will be added back to the supply as mintable tokens (up to a defined cap percentage in any given time period). These tokens that were burned (spent) are then re-minted and distributed to creators doing the relevant work. Once net new tokens enter the system via minting according to the minting schedule, the system will always maintain that number of circulating tokens. In the end state there will be 10 billion tokens in the system circulating in perpetuity.
As consumers burn the token, there is positive price pressure on the token; as tokens are minted, there is negative price pressure on the token. An equilibrium price is established based on these system dynamics.
In each period, the number of tokens minted and paid to contributors will depend on two factors. The first is the minting schedule of net new tokens released into the system as progress is made on the map. The second is the number of tokens bought and burned by customers in order to consume map data.
Net New Tokens Minted
Fig. 20 / Net New Tokens Minted Relative to Recirculated

Over time, the minting schedule dictates that fewer new tokens are introduced. However, the number of recirculated tokens due to the burn and mint equilibrium increases with customer demand.

Because of the burn and mint equilibrium, the network will continue to generate substantial rewards for contributors in perpetuity, and the amount of value available will eventually become tightly linked to the amount of value being paid for by customers, as the proportion of previously circulating tokens minted increases relative to the proportion of new tokens minted.
Hivemapper Token Economy
Fig. 21 / The Hivemapper Network

This diagram shows the two-sided marketplace between map contributors and map consumers interacting on the Hivemapper Network with its cryptocurrency token HONEY.

Token Restrictions for Employees and Investors
Tokens allocated to employees of Hivemapper Inc. are restricted for a three-year period with transfer restrictions released from Tokens in 24 equal tranches at the end of the 13th month to the end of the 36th month anniversary of the Initial Token Distribution Date.
Tokens allocated to the Investors are restricted for a two-year period with transfer restrictions released from Tokens in 12 equal tranches at the end of the 13th month to the end of the 24th month anniversary of the Initial Token Distribution Date.
The Initial Token Distribution Date is the day at which the token launches on mainnet.

Future Areas

Today, the mapping network focuses on street level maps. Next, the network will introduce map annotations enabling map editors to edit the map alongside the machine learning algorithms to add new layers of data to the map.
Yet, this is just the beginning. Global data collection for a global map, incentivized by the HONEY cryptocurrency token and protocol, is the right approach to building a global map. With this in mind, there are multiple technologies that hardware and software developers can use to add additional layers to the global map in order to meet the needs of customers:
  • Additional imagery sensors such as 360 cameras from street level10
  • Support for collecting imagery from scooters and bikes to support high quality maps for alternative transportation modalities
  • Use of air quality sensors to incorporate air quality data into the map11
  • Use of lower cost RGB-D, radar, and LiDAR sensors to build 3D maps and street level object mapping
  • Airborne data collection via drones to provide the high precision aerial perspective12
  • Use of satellite imagery for a broad scale aerial perspective

Notes & References

Our estimate of $300 billion includes mapping APIs, GIS, Geo Analytics, and data collection services.
Google consistently raises prices for mapping APIs, and 4 to 5 million developers use Google Maps APIs in their apps.
The Open Dashcam's hardware and software is open source. The firmware is written in Zig.
HONEY tokens will be distributed in the testnet phase exclusively for testing purposes. These tokens cannot be used or sold outside of the network. To avoid any confusion, these tokens will carry a modified name.
Once the Hivemapper mapping network is live on the blockchain, the HONEY token is the only form of reward. Contributions made prior to the launch of the Hivemapper blockchain mainnet beta will not be added to the blockchain.
The Hivemapper Foundation will be established as a US based 501(c)(6) non-profit entity. No Hivemapper Inc. executives will hold positions in the Hivemapper Foundation entity.
We define the progress for some region for some week as a function of that regions’ weekly:
Coverage is defined as the proportion of target L12 H3 cells covered. The set of target L12 H3 cells is determined by the foundation — for now it is generated via OSM road network data.
Today, high quality 360 cameras needed for mapping begin at ~$15,000. The Insta 360 Titan is one such example. Hardware partners who can help reduce these costs by a few orders of magnitude would be of interest to the mapping network.
To build an air quality sensor into the Open Dashcam it must be mounted externally and remain fixed to the vehicle.
Our attempts at building a network of drone mappers to build large scale aerial maps demonstrated that it's currently not economically viable. A commercial drone must fly for many hours and beyond the line of sight to make large scale drone mapping economically viable for contributors and consumers.