Keep the first experiment smaller than the ambition
NODERIQ studies how an autonomous task remains understandable when devices, networks and a supervising workflow no longer share the same state. The first proposed unit is one interrupted handoff: a task is issued, an action may have happened, and the completion signal is delayed or lost. The test must distinguish a confirmed result from an uncertain one before it permits another command.
Recent project exchanges have narrowed the planning around this bounded question. Candidate contributions span replayable software scenarios, controlled network impairment, device-side observations and recovery review. Those scopes are still being matched to an experiment, resources and eligibility; they are not a confirmed consortium or a claim of working cross-system integration.
Move only when the previous stage leaves a usable record
The proposed sequence begins in software. A recorded work order is replayed with a known baseline, then with a delayed acknowledgement, dropped connection or conflicting observation. The expected output is a timeline of intent, command, acknowledgement and physical-outcome evidence, including an explicit unresolved state when proof is missing.
A controlled laboratory stage would repeat the same task under measured network changes and compare the task record with observations from the device boundary. A physical or operational setting is a later possibility, dependent on access, equipment, site rules, safety review and a separately approved test plan. Device-local motion and emergency controls remain authoritative throughout.
- Define one work order, success condition, stop condition and classical baseline.
- Inject failure and replay the trace; classify outcomes as complete, uncertain or stopped.
- Advance to a controlled device test only when the interfaces and evidence can be checked.
- Consider a site trial only after equipment, access and operational approval are secured.
The proposal has its own acceptance gates
The current CELTIC-NEXT call is one possible application route. Before a formal submission, the project needs an agreed scope, a credible experiment, resourced contributions and country-specific funding eligibility. An alternative call can be considered if those conditions cannot be met in time; each route has its own rules and review. Circulating a presentation or discussing a contribution is not an application decision.
No funded project, signed participation, authorized physical pilot or measured field result is being announced. The next useful artefact is a test plan that states what would count as task completion, what remains uncertain after reconnection and who can release a recovery action without giving software authority over local safety.
BA · Bulletin annexDefinitions, recurring questions, source notes, and publication tags.Inspect
SNSource notes
3 recordsTags
