Feedback Loops · GBB Embodied Interaction AI Institute

Research question

An action ending does not mean the task is complete. Execution acknowledgments, environmental changes and human judgments may disagree. Distinguish commands sent, devices responding and goals achieved.

Research method

Build an event trace that records input states, candidate plans, action requests, execution acknowledgments and observed outcomes separately. Preserve the last trusted state after failure and specify when to retry, roll back or request human control.

Validation approach

Create counterexamples where the command succeeds but the goal is not achieved.
Simulate a disconnected device and check for false completion reports.
Check whether the records reconstruct decision evidence and interruption points.

Discuss a research question

RESEARCH PROTOCOL

Feedback loops: turning failure into the next input

Retain the pre-action state, expected result, actual observation and termination reason for every execution. Classify failures as perception errors, unreachable plans, execution deviations or external intervention. Use that distinction to choose a new observation, local replanning or human assistance.

Experimental design & observations

Stage	What to record	What to check
Deviation detection	Expected position, actual position, tolerance	Can noise be distinguished from a real deviation?
Failure recovery	Failure type, available alternative actions	Does it repeat the same ineffective action?
Evidence replay	Input version, action sequence, event time	Can the original decision evidence be reconstructed?

Making results comparable

Fix the initial map, endpoints and input version, then change one environmental condition at a time. Record the baseline before introducing obstacles or conflicting inputs. Include successful paths and failure cases. Compare strategies using path length, explored cells and execution traces.

Reproducibility checklist

Input and map snapshots: record scene versions, task conditions and random seeds.
Strategy configuration: record algorithms, constraints, timeouts and human intervention points.
Execution traces: retain state changes, action sequences, failure causes and recovery steps.
Comparison notes: document variable changes, observations and follow-up validation.