uid stringclasses 10
values | dataset_id stringclasses 1
value | year int64 2.02k 2.02k | setting stringclasses 5
values | task stringclasses 5
values | perception_detected_target stringclasses 1
value | pose_estimate_quality stringclasses 3
values | grasp_or_action_plan_valid stringclasses 2
values | collision_check_passed stringclasses 2
values | execution_completed stringclasses 2
values | observed_outcome_matches_intent stringclasses 2
values | ops_summary stringclasses 10
values | question stringclasses 1
value | ground_truth_label stringclasses 2
values | ground_truth_rationale stringclasses 10
values | outcome_signal stringclasses 4
values | source_citation stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
RBPA-001 | robotics-perception-action-coherence-risk-v0.1 | 2,024 | warehouse | pick_place | yes | high | yes | yes | yes | yes | Target detected, plan valid, execution matches intent. | Is perception-action coherent. Answer coherent or incoherent. | coherent | Perception and action remain aligned. | successful_task | Synthetic |
RBPA-002 | robotics-perception-action-coherence-risk-v0.1 | 2,024 | warehouse | pick_place | yes | low | no | no | no | no | Target detected but pose poor, plan fails, execution aborts. | Is perception-action coherent. Answer coherent or incoherent. | incoherent | Perception uncertainty breaks action. | failure_risk | Synthetic |
RBPA-003 | robotics-perception-action-coherence-risk-v0.1 | 2,023 | factory | bin_picking | yes | high | yes | yes | yes | yes | Accurate pose leads to clean grasp and lift. | Is perception-action coherent. Answer coherent or incoherent. | coherent | Action matches perception. | successful_task | Synthetic |
RBPA-004 | robotics-perception-action-coherence-risk-v0.1 | 2,023 | factory | bin_picking | yes | medium | yes | yes | yes | no | Robot grasps wrong item despite detection. | Is perception-action coherent. Answer coherent or incoherent. | incoherent | Outcome diverges from intent. | quality_escape | Synthetic |
RBPA-005 | robotics-perception-action-coherence-risk-v0.1 | 2,022 | lab | door_opening | yes | high | yes | yes | yes | yes | Handle detected, trajectory executed, door opens. | Is perception-action coherent. Answer coherent or incoherent. | coherent | Intent achieved. | successful_task | Synthetic |
RBPA-006 | robotics-perception-action-coherence-risk-v0.1 | 2,022 | lab | door_opening | yes | high | yes | no | no | no | Handle detected but collision check fails, action unsafe. | Is perception-action coherent. Answer coherent or incoherent. | incoherent | Safety check breaks execution. | safety_risk | Synthetic |
RBPA-007 | robotics-perception-action-coherence-risk-v0.1 | 2,021 | hospital | delivery | yes | medium | yes | yes | yes | yes | Robot navigates to target and delivers payload. | Is perception-action coherent. Answer coherent or incoherent. | coherent | Perception supports correct action. | successful_task | Synthetic |
RBPA-008 | robotics-perception-action-coherence-risk-v0.1 | 2,021 | hospital | delivery | yes | low | no | no | no | no | Robot mis-localises and stops in wrong bay. | Is perception-action coherent. Answer coherent or incoherent. | incoherent | Perception drift prevents completion. | failure_risk | Synthetic |
RBPA-009 | robotics-perception-action-coherence-risk-v0.1 | 2,020 | field | inspection | yes | high | yes | yes | yes | yes | Object detected and inspected with correct standoff. | Is perception-action coherent. Answer coherent or incoherent. | coherent | Tracking and control aligned. | successful_task | Synthetic |
RBPA-010 | robotics-perception-action-coherence-risk-v0.1 | 2,020 | field | inspection | yes | medium | yes | yes | yes | no | Approach overshoots and loses target. | Is perception-action coherent. Answer coherent or incoherent. | incoherent | Control does not realise perceived plan. | failure_risk | Synthetic |
What this repo is for
You use it to detect when a robot sees the right thing but does the wrong thing.
It targets day-to-day deployment failures:
correct detection with wrong grasp
pose uncertainty that breaks planning
safety check failure that blocks execution
execution that completes but misses intent
Typical uses:
warehouse pick and place
bin picking
hospital delivery robots
field inspection robots
Prompt format
Output must be exactly one token
coherent or incoherent
- Downloads last month
- 21