Skip to content

Supporting subject

Networks and DePIN

Decentralized physical infrastructure means many people deploy hardware, software checks that the hardware did a useful job, and payment follows that job. Wireless coverage, positioning, sensing, storage and compute are the usual services. A token with no machine is not part of this section.

Services, with examples

Examples name the kind of network, not a recommendation. Projects shut down, rename and change what they sell. Inclusion is not an endorsement.

  • Position

    Make outdoor location sharp enough for machines, not just for turn-by-turn maps.

    GEODNET is the clearest public example of a decentralized GNSS correction network.

  • Identity and settlement

    Give a machine a durable ID, a record, and a way to pay or be paid across parties who do not share a login.

    peaq is the project this desk follows most closely for machine IDs and the machine-economy stack.

  • Shared space

    A map of rooms, venues and streets that many robots can use, instead of one private point cloud.

    Auki is an example of machine perception treated as a network.

  • Nearby compute

    Inference and other jobs when the robot’s own computer is too small or the link to a hyperscaler is the wrong shape.

    Acurast (phones as processors) and broader GPU networks sit in this aisle.

  • Wireless

    Coverage in places a single carrier buildout does not prioritize.

    Helium is the familiar pattern: hotspots as the infrastructure.

  • Mapping and sensing

    Street-level imagery, audio, weather and other samples a model cannot invent from text.

    NATIX-style drive-and-map networks, and audio or environment networks in the same family.

  • Device coordination

    Connecting machines and proving device activity as a base layer, not as an app feature.

    IoTeX is often discussed in this device-and-DePIN lane.

  • Collective machines

    Groups pooling capital around robots, with the hard part being whether revenue follows the machine.

    XMAQUINA is an example of a robotics-focused collective. Read the audit trail, not the adjective.

Five checks

Use these before a hardware network earns a closer look. They will not tell you what to buy. They will tell you whether a machine is actually in the story.

  1. 01What physical job is being done, in one sentence a technician would recognize?
  2. 02Who can check that the job happened, and which parts can be faked from a laptop?
  3. 03Who pays, and for which unit — not for “supporting the network”?
  4. 04What still works the week the reward price is boring and the power bill is not?
  5. 05Would a robot or an AI system buy it, or only another token holder?

Latest

Background