Categorization
There are many ways to categorize DePINs. We focus on the demand-side service provided by each network, because we believe that is the primary driver of revenue quality and scale over the long-term. Note that projects can switch categories over time as their demand-side focus shifts, and two projects with similar supply-side profiles can fall into different buckets.
We recognize eight categories of DePINs:
| Demand-side: Job-to-be-Done | |
|---|---|
| Sensors | Capture data |
| Compute | Store and manipulate data |
| Wireless | Move data from one location to another |
| Mobility | Move assets from one location to another |
| Energy | Generate and store energy |
| AI | Generate and store machine intelligence |
| Human Capital | Store human identity and reputation |
| Robotics | Provide offchain services delivered by machines |
Decentralized Compute networks, or DeVINs (Decentralized Virtual Infrastructure Networks), aggregate computing resources to provide memory, storage and compute capacity on-demand. DeVINs can be general-purpose or specialize into certain use cases or types of resources. DeVIN is the only category whose core service can be verified cryptographically, leading some investors to categorize them outside of DePIN entirely.
Decentralized Wireless networks, or DeWis, aggregate wireless resources like radios, antennas, satellites and IP addresses in order to provide bandwidth on-demand. DeWis generally leverage established wireless standards like WiFi, 4G/LTE, LoRa or Bluetooth in order to interoperate with existing device ecosystems. This category also includes privacy-focused routing or messaging protocols whose primary function is to move data privately between users in different locations.
Decentralized Energy networks, or DeGENs (Decentralized Generative Energy Networks), aggregate renewable energy resources like solar panels, batteries, electric vehicles and thermostats to provide electricity and/or carbon mitigation assets on-demand. Most DeGENs describe themselves as either decentralized virtual power plants and/or as self-perpetuating onchain sinks for carbon mitigation assets.
Decentralized Sensor networks generate novel datasets and provide access to the data and/or the insights derived from it. These networks can collect various types of data (e.g. text, video, audio, location), across a variety domains (e.g. weather, mobility, health, social media), and sell to various end-markets (e.g., web2 SaaS & marketplaces, AI developers, financial institutions).
Decentralized Mobility networks coordinate the logistics of moving physical assets like people (ride-sharing), consumables (food delivery), vehicles (mobility) and shipments (logistics) from one location to another. These networks generally combine software- and hardware-based verification to coordinate trips and mediate disputes in a trust-minimized manner.
Decentralized AI networks aggregate the various resources above - e.g. compute, energy, bandwidth and sensor data - to generate and provide access to machine intelligence. While each underlying resource contributes to the process, only networks that directly incentivize the development of machine intelligence fall into this category.
Decentralized Human Capital networks aggregate information about people’s identities and reputations in order to facilitate low-trust connections. This category includes general-purpose identity networks for proof-of-humanity; decentralized value-added marketplaces for labor, dating or social media influence; as well as onchain loyalty and rewards ecosystems.
Decentralized Robotics networks aggregate machines in order to provide offchain services like manufacturing, hospitality, construction, medical care and defense. This is currently the most nascent category of DePIN in terms of developer and investor activity.