Robot-As-A-Sensor: Forming a Sensing Network with Robots for Underground Mining Missions

TL;DR

The paper addresses the need for reliable, scalable collaborative sensing in underground mining by introducing Robot-As-A-Sensor (RAAS), a network of mobile robotic sensing units that form a local sensing fabric. It surveys sensing modalities and Wireless Sensor Network technologies, formalises RAAS, and analyzes challenges in scalability, mobility, coverage, and latency, offering a technical outlook centered on Integrated Sensing and Communication (ISAC) and edge computing. The key contributions include a conceptual RAAS architecture, a comparison with traditional WSNs, and proposed enabling technologies such as ISAC, EC, and semantic communications to realize robust, low-latency underground sensing networks. The work envisions Mining 5.0 where robotic sensing networks enhance safety, efficiency, and resilience through autonomous collaboration and advanced communication strategies.

Abstract

Nowadays, robots are deployed as mobile platforms equipped with sensing, communication and computing capabilities, especially in the mining industry, where they perform tasks in hazardous and repetitive environments. Despite their potential, individual robots face significant limitations when completing complex tasks that require the collaboration of multiple robots. This collaboration requires a robust wireless network to ensure operational efficiency and reliability. This paper introduces the concept of "Robot-As-A-Sensor" (RAAS), which treats the robots as mobile sensors within structures similar to Wireless Sensor Networks (WSNs). We later identify specific challenges in integrating RAAS technology and propose technological advancements to address these challenges. Finally, we provide an outlook about the technologies that can contribute to realising RAAS, suggesting that this approach could catalyse a shift towards safer, more intelligent, and sustainable industry practices. We believe that this innovative RAAS framework could significantly transform industries requiring advanced technological integration.

Figures (2)

• Figure 1: A graphical illustration of Robot-As-A-Sensor for underground mine sites. The dashed lines indicate the paths and patterns that the robots follow during inspections. In a dynamic environment, robots can adapt these paths based on real-time data collected from other robots or their own sensors in the operation area. The solid arrows represent the high bandwidth communication channels established between the robots, which can be used for transmitting multimedia data. The dotted arrows represent low-bandwidth communications that might be used for telegraphing sensor data and/or control signals. The green shaded area depicts the area where the backhaul network is accessible. Warnings for Hazards and Sources of Hazards Critical for automated inspections in potentially hazardous environments. Robots can detect and communicate hazards to each other, enhancing situational awareness and safety.
• Figure 2: An illustration of one of the possible architectures of RAAS Networks. The illustration shows that an RAAS network effectively integrates multiple RAAS members to perform specific tasks and share data crucial for operational success in challenging underground environments. For each RAAS member within the network (e.g., RAAS Member $i$), it is capable of conducting multi-mode sensing and communications between neighbouring members and dynamically controlling the progress of existing missions, which is particularly designed to operate within the complex and hazardous confines of an underground mine.