feedoracle.io Contact
  1. Home
  2. Technologies
  3. AI & Autonomous Systems
  4. Edge AI & IoT Research

Edge AI & IoT Research

Intelligence where things happen, not only where data ends up.

A research area within AI & Autonomous Systems. Our current work is an evaluation-stage internal prototype (EDGE-01) for governance that keeps working when a device is offline. Until November 2026 this page presented IoT & Edge as a separate technology field.

Edge node decides locallyqueued while offlineremote recordsynced on reconnectResearch concept, not deployed hardware

Research status

This area is in research. The following is not established, and we do not claim it.

  • No physical device testsThe prototype was evaluated in a synthetic setting, not on physical edge hardware such as a Jetson board. Offline operation on real devices is not demonstrated.
  • No production deploymentThere is no IoT deployment in production.
  • No hardware security certificationNo device or component has been certified.

The domain

Decisions close to the device, and records that survive a lost connection.

At the edge, the network is not always there. Decisions and records have to work without it.

Connected devices and edge computers sense and act where things happen. Sending every signal to a central system is slow, costly and fails when the connection drops.

Our research looks at how an edge device can decide whether an action may run, keep a signed record of it while offline, and reconcile that record once it is connected again.

Research questions

Where this field stands: questions first, capabilities second.

  • Which decisions can an edge device make on its own, and which need a remote check?
  • How does a record stay trustworthy while a device is offline?
  • How can sensor data keep its origin and its age from the device to the decision?

Example: offline-capable governance

How the research prototype handles an action without a connection, and what happens when it returns.

Offline

  1. Request on the device

    An action is requested while the device is offline.

  2. Local authorization

    Checked against the policy available on the device.

  3. Run or refuse

    The action runs only if it was permitted.

  4. Receipt signed locally

    A record of what ran, signed on the device.

  5. Queued

    The receipt waits for a connection.

Reconnected

  1. Queue sent upstream

    Queued receipts are sent in order.

  2. Receipts checked

    Signatures and order are checked remotely.

  3. Records reconciled

    The remote record now matches what ran at the edge.

In research An internal prototype in a synthetic technical evaluation. It is not a product, it has not been tested on physical devices, and no evaluation material is published.

Capabilities and status

Research work and focus areas. No capability in this area is shown as implemented or demonstrated.

Demonstratedtested capability or reproducible public demonstrationImplemented / documentedexists and is documented, current function not independently verifiedIn researchexperimental or exploratory workPlannedplanned or staged, not activeStatus to confirmnot enough evidence to classify

Offline-capable governance

An internal research prototype, evaluated in a synthetic technical setting.

  • Offline authorization

    Deciding whether an action may run while the device has no connection.

    In researchResearch prototype
  • Locally signed receipts

    Records of what ran, signed on the device itself.

    In researchResearch prototype
  • Queued recovery and synchronization

    Records wait in a queue and are synchronized and checked once the connection returns.

    In researchResearch prototype

Connected devices and machine communication

Devices that talk to each other and to services.

  • Device integration

    No production integration. Not enough evidence yet to classify further work.

    Status to confirm—
  • Machine-to-machine communication

    Named as a focus area. Not enough evidence yet to classify.

    Status to confirm—

Edge computing and sensor intelligence

Computation close to where signals arise.

  • Edge computing

    Named as a focus area. Not enough evidence yet to classify.

    Status to confirm—
  • Sensor intelligence

    Named as a focus area. Not enough evidence yet to classify.

    Status to confirm—
  • Autonomous physical systems

    Named as a focus area. Not enough evidence yet to classify.

    Status to confirm—

Developments in this area

None yet.

There is no development in this area yet. The prototype above is research and stays in the Research Center until it is demonstrated on real devices.

Connections to the technology fields

Where edge research meets the technology fields.

  • AI & Autonomous Systems

    Governance for autonomous actions that keeps working offline.

    In research
  • Data & Intelligence

    Sensor data that keeps its origin and its age.

    Status to confirm
  • Blockchain & Distributed Infrastructure

    Machines paying each other for data or services.

    Status to confirm

Relationship to the Research Center

This area is research today. It feeds two research directions, and its prototype lives in the Research Center until it is demonstrated on real devices.

FeedOracle Research Center
  • Autonomous systems research
  • Novel computational architectures

Work with us

Real starting points only. Pages that do not exist yet are not linked.

  • Research collaboration

    The main way into this area today. Bring a question about edge governance, offline operation or sensor data.

  • Developers

    There is no developer interface for this field yet. Related governance work is documented for AgentNOMOS.