Research agenda

Roadmap

Twelve months, grouped by program. Status comes from the lab data file. No journal names.

Timeline

AI-Pass submission
In progress
Quality orchestration charter and shared invoice benchmark
Planned
Simulation baseline with fault injection (ROS 2, Gazebo)
Planned
Graph-pruning study
Planned
First quality-routing results
Planned
Sovra raw data and scoping
In progress
Degradation curves and verification methodology note
Planned
Integrated pipeline paper
Planned
Open benchmark for extraction or orchestration under constraint
Planned
View as table
Roadmap
ItemProgramMonthsStatus
AI-Pass submissionroute0 to 2In progress
Quality orchestration charter and shared invoice benchmarkassure0 to 3Planned
Simulation baseline with fault injection (ROS 2, Gazebo)physical0 to 3Planned
Graph-pruning studyground3 to 6Planned
First quality-routing resultsassure3 to 6Planned
Sovra raw data and scopingfoundation0 to 3In progress
Degradation curves and verification methodology notephysical3 to 6Planned
Integrated pipeline paperroute6 to 12Planned
Open benchmark for extraction or orchestration under constraintfoundation6 to 12Planned

Publication stage gates

Scoping, logs, draft, internal review, then submission.

No items in the publication pipeline yet.

Scoping0
Data and logs reviewed0
Draft0
Internal referee-style review0
Submission0

Planned artifact releases

  • Quality orchestration charter (Planned)
  • Shared invoice testbed protocol (Planned)
  • Graph-pruning study protocol (Planned)
  • Sovra scoping note and raw-log schema (In progress)

Qualitative risks

low / high
Safety overclaim
Dual-use drift
medium / high
Dependence on student labor
Benchmark credibility
Overclaimed service levels
high / high
low / medium
medium / medium
high / medium
Spread across too many tracks
low / low
medium / low
high / low
View as table
Risks
RiskLikelihoodImpactMitigation
Dependence on student labormediumhighTwo people per program, documented pipelines
Spread across too many trackshighmediumOne lead per track, small physical track
Benchmark credibilitymediumhighIndependent adjudication, disclosed construction
Overclaimed service levelsmediumhighState assumptions explicitly
Safety overclaimlowhighName the testbed stage for every claim
Dual-use driftlowhighWritten policy before any public release