SAFARI: Semantic- And Frontier-driven Autonomous Risk Inspection
SAFARI: Semantic- And Frontier-driven Autonomous Risk Inspection
This paper introduces SAFARI, a framework that unifies semantic object detection with frontier-based exploration to enable risk-aware autonomous mapping in hazardous environments. Evaluated across 11 missions on a quadruped platform at three field sites, the system achieved 98.7% coverage of accessible areas and operated without human intervention for 97.6% of mission time, detecting objects as small as 15cm. Ablation studies highlight the importance of object-size awareness and safety-aware path planning for effective hazardous-object search.
The authors present SAFARI, a framework integrating semantic object detection and risk assessment into autonomous exploration on a quadruped platform, validated across 11 field missions with reported metrics including 98.7% area coverage and 97.6% operator-free operation time.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Control
- Reported autonomousOpen source S2
Reported numbers
minimum_detected_object_size_cm
15
View original evidence
tion and planning with both geometric and semantic evidence. We evaluated SAFARI in 11 missions across three field sites on a quadruped platform. The framework achieved an average coverage of 98.7% of the accessible mission area, without operator intervention for 97.6% of mission time, and detected objects with characteristic dimensions as small as 15cm. Lastly, through ablation studies and a detailed analysis of sys
Open source S4mission_count
11
Reported trials
View original evidence
tion and planning with both geometric and semantic evidence. We evaluated SAFARI in 11 missions across three field sites on a quadruped platform. The framework achieved an average coverage of 98.7% of the accessible mission area, without operator intervention for 97.6% of mission time, and detected objects with characteristic dimensions as small as 15cm. Lastly, through ablation studies and a detailed analysis of sys
Open source S4field_sites_count
3
Reported trials
View original evidence
tion and planning with both geometric and semantic evidence. We evaluated SAFARI in 11 missions across three field sites on a quadruped platform. The framework achieved an average coverage of 98.7% of the accessible mission area, without operator intervention for 97.6% of mission time, and detected objects with characteristic dimensions as small as 15cm. Lastly, through ablation studies and a detailed analysis of sys
Open source S4
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
d with operating near potentially hazardous objects. Moreover, many systems have been demonstrated primarily in simulation or controlled environments, leaving limited validation of risk-aware semantic exploration in complex, unstructured field settings. To address this gap, we propose SAFARI, a framework that treats object detection as a native channel of a unified probabilistic map, thereby jointly informing explora
Open source S3
arXiv:2610.41456v1 Announce Type: new Abstract: Robotic exploration can reduce human exposure to hazardous environments by autonomously searching unknown areas for objects of interest while constructing a complete, risk-aware map. However, existing autonomous exploration approaches typically prioritize geometric coverage or information gain, without explicitly coupling semantic object detection and the risk associate
Open source S2
tion and planning with both geometric and semantic evidence. We evaluated SAFARI in 11 missions across three field sites on a quadruped platform. The framework achieved an average coverage of 98.7% of the accessible mission area, without operator intervention for 97.6% of mission time, and detected objects with characteristic dimensions as small as 15cm. Lastly, through ablation studies and a detailed analysis of sys
Open source S4
Source：arXiv Robotics — research abstracts · arxiv.org