Improving mineworker safety using a mineworker administration shell

Despite rapid technological advancements in automated mining machinery, underground mineworkers remain highly vulnerable to severe environmental hazards, operating within fragmented safety systems that treat them as passive data subjects. This article introduces the mineworker administration shell (M-AS), a decentralized digital integration mechanism that transforms the worker into an active, coordinated component of a human cyber-physical system. By dynamically connecting environmental sensors with central mine information infrastructures, the M-AS enables real-time hazard detection, automated personalized evacuation routing and proactive, data-driven pre- and post-shift health monitoring. This framework aims to leverage existing data streams to eliminate generic one-size-fits-all safety protocols, directly lowering injury rates and health issues, and ensuring that no worker is allocated to an unsafe environment or shift for which they are physically unfit. Background Modern underground mining relies heavily on human labor, yet dangerous working conditions persist. For example, the South African mining industry alone reported more than 2,000 injuries and 74 fatalities in underground environments in 2021. While Industry 4.0 innovations have introduced wearable safety devices, like smart helmets and internet-of-things (IoT) jackets, these solutions typically function as stand-alone, isolated data loggers. They generate data silos that fail to achieve proactive safety management or systemic integration with wider mine information and control networks. Furthermore, underground environments introduce severe operational constraints, such as network signal reflections, a lack of worker position monitoring in the mine, and microclimate variations that degrade low-cost sensor accuracy, which limit the scale of traditional digital twin applications. To bridge this gap, an integrated cyber-physical architecture is required to actively represent and coordinate mineworkers within the primary operational loop. Methods Fig. 1 M-AS components using BASE architecture. The M-AS was conceptually developed using the biography-attributes-schedule-execution (BASE) architecture to manage data across distinct temporal boundaries. The framework consists of four core components: (1) Biography (B): A digital archive storing long-term, historical data on completed tasks and past environmental exposures. (2) Attributes (A): A repository for static or slow-changing administrative metrics and personalized health thresholds. (3) Schedule (S): A module managing data related to future planned shifts and coordinated task allocations. (4) Execution (E): A real-time data space hosting a state blackboard to process highly volatile, in-shift biometric and ambient measurements.Application-specific plugins execute the data acquisition and analysis functions that extend this architecture to evaluate distinct risks. Figure 1 shows the various components and plugins used within each M-AS using the BASE architecture. For heat stress monitoring, a scheduling plugin adds monitoring intervals to the schedule, while an execution plugin captures dry-bulb temperature, TDB, wet-bulb temperature, TWB, relative humidity, RH, and air velocity, υ. The data are used to dynamically calculate the thermal humidity index, THI, and wet-bulb globe temperature, WBGT: THI = TDB – 0.55 × (1 – RH) × (TDB – 14.5)WBGT = 0.3TDB + 0.7TWB For gas hazard monitoring, plugins interact with handheld gas detection instruments to capture concentrations of particulate matter, sulfur dioxide (SO2), nitrogen oxides (NOx), methane (CH4) and carbon monoxide (CO). These inputs generate a comprehensive scale index (CSI) to classify ambient air quality into explicit risk tiers.The system was implemented using the functional programming language Erlang via the BASE platform and integrated into a simulated underground mine information system (UMIS). Figure 2 shows the architecture of the UMIS. Testing was performed within the smart integrated mining laboratory (SIMLAB) at Stellenbosch University using rugged Android tablets connected over replicated underground network infrastructure. Results The M-AS framework was verified through two targeted laboratory experiments involving three mineworkers and one miner over accelerated eight-hour shift cycles. In the context of the following case study, a miner is an employee with managerial responsibilities and specialized safety equipment like a gas detection instrument, whereas a mineworker refers to an employee who performs the most physically demanding work under the most arduous underground conditions. Fig. 2 Underground mine information system. Normal operation experiment. During standard shift activities, execution plugins successfully requested and processed environmental variables at simulated five-minute intervals. Real-time metrics were compared continuously against predefined thresholds, updating individual worker user interfaces (UIs) via a green-amber-red traffic-light indicator. Upon shift completion, reflection plugins curated raw execution variables into long-term cumulative averages within the biography component. The analysis plugin automatically processed these historical blocks to update each worker’s “activity readiness” attribute ahead of the subsequent shift loop. Performance logs indicated highly efficient software execution, with a mean application CPU load of 4.15 percent and average RAM usage of 5 MB. UI latency for key operations remained low, as detailed in Table 1. Safety incident experiment. An emergency scenario was introduced by simulating a localized gas leak in an active workplace. Upon receiving an automated hazard alert, the safety incident coordination (SIC) process within the UMIS bypassed standard task scheduling and initiated an active worker evacuation protocol via the target M-AS instances. The system successfully distributed distinct, prioritized evacuation routes based on localized worker positioning data. The progression of warning states and automated spatial transitions across the network is recorded in Table 2. Discussion The findings demonstrate that the M-AS effectively accommodates both reactive and proactive safety functions within a single integration architecture. In-shift monitoring functions act as an immediate reactive mechanism to alert workers as environmental safety levels decay due to the simulated gas leak. Concurrently, post-shift analysis acts as a proactive gatekeeper to prevent long-term exposure accumulation, replacing uniform, rigid industry rules with individualized safety tracking. Furthermore, the integration model supports distinct operational scales of personalization and specialization. Specialization allows the architecture to alter its interface functionality based on explicit underground roles (such as distinguishing a miner’s supervisory view from a mineworker’s task view), while personalization relies on unique attribute configurations to maintain an accurate digital representation of individual human tolerances.However, system limitations remain. The accelerated timeframes used in laboratory simulation (96 seconds representing an eight-hour shift) served to validate software logic but could not capture long-term physical microclimate variations, such as the thermal inertia of a rock face. Practical field implementation will require localized sensor array calibration to account for complex ventilation dynamics. Conclusions The fragmentation of environmental monitoring technologies limits holistic safety management in underground mines. By deploying M-ASs built on the modular BASE architecture, the mining industry can successfully transform individual operators into active, highly integrated assets within the smart mine information loop. This approach optimizes human resource utilization and drastically accelerates emergency response times during critical incidents. Future research trajectories must pursue a phased deployment strategy, initiating field trials in nonproduction training stopes to benchmark network resilience under real operational conditions and evaluate broader worker usability and technology acceptance. n References A list of all references is available in the full paper.

Improving mineworker safety using a mineworker administration shell

Authored by: Sa’eed Fataar, Karel Kruger, Anton Herman Basson and Nicole Catherine Taylor

To view this article you must be a member of SME/UCA.

US Energy Department launches $16 million prize to grow the mining and critical minerals workforce

DOE selects S&T to build mining technology proving ground

Legacy in Motion: How CAB is Building the Future of Hanger Manufacturing

US Energy Department launches $16 million prize to grow the mining and critical minerals workforce

DOE selects S&T to build mining technology proving ground

Legacy in Motion: How CAB is Building the Future of Hanger Manufacturing

To view this article you must be a member of SME/UCA.

Highland Copper announces final approval of US$50 million grant front the state of Michigan for Copperwood Project

A greedy genetic hybrid heuristic for solving the openpit production scheduling problem

Related Articles

A greedy genetic hybrid heuristic for solving the openpit production scheduling problem

Application of infrared thermography and the acoustic emission technology for loose rock detection

Enhancing quartz-amine flotation with oily bubbles

Rheological characterization of tailing pastes from ore mining

US Energy Department launches $16 million prize to grow the mining and critical minerals workforce

DOE selects S&T to build mining technology proving ground

Legacy in Motion: How CAB is Building the Future of Hanger Manufacturing

US Energy Department launches $16 million prize to grow the mining and critical minerals workforce

DOE selects S&T to build mining technology proving ground

Legacy in Motion: How CAB is Building the Future of Hanger Manufacturing

Use Single Sign On