SCALOFT: An Initial Approach for Situation Coverage-Based Safety Analysis of an Autonomous Aerial Drone in a Mine Environment

Nawshin Mannan Proma, Victoria J Hodge, Rob Alexander · 2025

Read SCALOFT: An Initial Approach for Situation Coverage-Based Safety Analysis of an Autonomous Aerial Drone in a Mine Environment by Nawshin Mannan Proma, Victoria J Hodge, Rob Alexander (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-20969.txt.

Introduction

Proma, Hodge, and Alexander test a safety check frame for a self-flying drone in a mine in this twenty twenty-five paper. Mines are dark, tight, and risky, so a drone that flies itself must be checked hard. Their frame builds wide test cases and watches the drone live as it flies.

Inside, each breach caught in flight gets a code and a log line for later study. A score for how much of the scene space tests have touched tops off the frame. The team then plants known faults in the drone to grade the frame's eye. For a small set of fair faults, the frame caught each one as hoped. That backs its use in a safety case for the drone.

Catalog record

CREATORNawshin Mannan Proma, Victoria J Hodge, Rob Alexander
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE19 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-20969

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

SCALOFT: An Initial Approach for Situation Coverage-Based Safety Analysis of an Autonomous Aerial Drone in a Mine Environment Nawshin Mannan Proma, Victoria J Hodge, Rob Alexander Abstract The safety of autonomous systems in dynamic and hazardous environments poses significant challenges. This paper presents a testing approach named SCALOFT for systematically assessing the safety of an autonomous aerial drone in a mine. SCALOFT provides a framework for developing diverse test cases, real-time monitoring of system behaviour, and detection of safety violations. Detected violations are then logged with unique identifiers for detailed analysis and future improvement. SCALOFT helps build a safety argument by monitoring situation coverage and calculating a final coverage measure. We have evaluated the performance of this approach by deliberately introducing seeded faults into the system and assessing whether SCALOFT is able to detect those faults. For a small set of plausible faults, we show that SCALOFT is successful in this. 1 Introduction Autonomous aerial drones (AAD) have gained popularity due to their ability to operate autonomously with minimal or no human intervention thus minimising human risk and offering cost-effective solutions [ 25 ] . Their applications include search and rescue missions, building inspections and navigating challenging environments such as underground mines. However, the dynamic nature of these operational environments necessitates the development of a systematic testing approach to ensure the safety of both the drones and their operating contexts. Establishing a detailed safety assessment process during the design phase [ 10 ] of AAD applications is difficult, especially where humans will be present, such as in underground mines. Such a process would need to identify a representative set of potential failure situations during AAD operations, assess their consequences, and define mitigation measures to minimize risks. For example, even if a drone can fly safely in different light conditions, we still need to make sure it follows the safety requirements when it is flying in complicated situations and ensure the risk posed is as low as reasonably practicable. In traditional software testing, it is common to use coverage measures to check if all parts of the software were tested. This could be as simple as making sure every line of code was run at least once. However, these measures have been criticised for not capturing all the important aspects of the software [ 5 ] . They might miss things like external factors that can affect how the software behaves. Traditional safety testing employs a variety of coverage techniques, such as system coverage, requirement coverage, and scenario coverage (refer to Table 1 for an AAD example of how these differ from one another). But testing autonomous software is always tricky because AADs can encounter all sorts of different situations, such as obstacles or interactions with people, in a constantly changing environment. To test autonomous system software in both expected and unexpected situations, it is necessary to consider a sufficient range of situations while testing [ 5 , 7 ] . Table 1: Coverage based safety testing Testing Approach Focus Key Attributes Example of AAD Testing in Mine System Coverage [ 20 ] , [ 30 ] , [ 14 ] , [ 11 ] , [ 15 ] Thorough testing of system components and interactions Test all components and their interactions Ensuring sensors, decision-making algorithms, and other systems work together effectively to avoid collisions Requirement Coverage [ 5 ] Verifying compliance with specified safety requirements Focuses on fulfilling explicit requirements Ensuring the requirement that AAD slows within a specified distance when a person is detected inside the mine Scenario Coverage [ 31 ] , [ 3 ] , [ 13 ] , [ 17 ] , [ 18 ] Performance in predefined real-world scenarios Evaluates system behavior in a sequence of events Testing drone’s behavior at predefined waypoints in the mine Situation Coverage [ 5 ] , [ 27 ] , [ 21 ] , [ 22 ] , [ 16 ] , [ 6 ] System’s adaptability in dynamic, real-time conditions Assess system’s ability to handle unexpected situations Evaluating how the drone responds to random obstacles inside the mine Situation coverage-based safety testing [ 5 ] assesses the system’s performance in dynamic, real-time situations, testing its robustness and adaptability to new situations. While this approach effectively tests the system’s ability to handle unforeseen events, it is impossible to ensure exhaustive coverage of all possible situations in system-level testing. Some recent literature describes how to find representative situations for autonomous vehicles (AVs) using a situation-coverage-based approach [ 21 , 27 ] . Currently, the methods can only generate simple test situations. These might help find bugs, but they do not cover all possible situations that can occur in the real world. Also, while some research has been conducted on situation coverage testing for AVs [ 22 , 24 ] , there is a lack of comprehensive studies addressing situation generation and safety testing of AADs to address the unique challenges of underground environments. This gap highlights the need for dedicated research work to develop and validate a situation coverage-based safety testing approach that addresses the operational requirements of AADs in underground environments such as mines. Due to the identified research gaps, this study aims to explore the construction of a situation hyperspace — a conceptual model that systematically organizes various factors, including environmental conditions, system states, and external influences, to comprehensively represent the possible combinations of operating scenarios for AAD based on the Operational Domain Model (ODM). ODM defines the specific conditions under which an autonomous system is designed to function safely to establish the system’s operational boundaries (discussed next in section 2 ). Additionally, the research will assess whether the test cases derived from this situation hyperspace provide adequate coverage of the potential situations an AAD system may encounter. Before conducting situation coverage-based testing, this study assumes that component-level coverage has already been achieved and is sufficient. While ensuring component-level coverage is a significant challenge, it falls outside the scope of this research. 2 Operational Domain Model Fig. 1 shows a zoomed view of a portion of an ODM for an underground mine. The ODM offers a structured representation of operational scenarios, environmental conditions, domain-specific factors, and mine structures. The total number of possible test case combinations derived from the ODM is calculated by analysing the key categories and their branches. Figure 1: Zoomed view of ODM for an underground mine. See [ 9 ] for the full ODM For example, in the ODM, the Operational Time category combines factors such as time metrics, restrictions, and duration metrics. Within the Environment category , various factors are considered, including airflow, visibility, moisture, temperature, dust, light, and methane ( CH_{4} ), each branching into multiple subcategories. The Operational Domain category also includes three types of dynamic objects (Machinery, Pilot, and Crew) and static objects. Additionally, the Mine Structure category accounts for elements such as ceiling/roof, wall materials, and floor materials, each offering multiple options. The combination of these variables results in thousands of unique test case possibilities. While this systematic approach ensures comprehensive situation coverage, simulating every combination is often infeasible due to cost and resource constraints. Therefore, we first need to select which combinations to test. To do this we

Read the complete text