Mining Contract: Designing a Statistical Algorithm for Strategic Identification and Development of New Mine Safety Technologies and Technological Applications
To identify hazards in a timely fashion, NIOSH must survey the landscape of mine hazards through a lens that is wide enough to capture the diverse range of mine operations, fine-grained enough to pinpoint hazards that pertain to each specific type of operation, and responsive enough to detect new hazards as they emerge.
Contract Status & Impact
This contract is complete. To receive a copy of the final report, send a request to email@example.com.
Two characteristics of the mining industry make it difficult to know which technologies are most likely to improve the safety and health of miners. First, mining is highly heterogeneous. For example, the material being mined, scale of mining activity, presence of underground or surface operations, timing of work shifts, geological conditions, technologies employed, ownership structure, and workforce characteristics can all vary widely, even within a single mining sector. Given this extraordinary diversity, identifying which categories of mining equipment are the most likely to place miners in harm’s way poses enormous informational challenges. Secondly, the mining sector is highly dynamic. Dissemination of new technology and equipment, fluctuating market conditions, and changes in regulatory oversight can rapidly alter the profile of occupational risks in some mines while leaving others unaffected. Therefore, even if NIOSH effectively promotes new technologies to mitigate well-known equipment hazards, other equipment problems that pose equally severe occupational risks may go undetected for some time.
The Mine Safety and Health Administration (MSHA) inspects each US mine at least twice (and for underground mines, at least four times) per year. In a previous contract, MSHA’s Part 50 data was used—in combination with MSHA’s internal enforcement data and several data elements obtained from NIOSH and the Department of Energy’s Energy Information Administration (EIA)—to produce an algorithm that predicted each coal mine’s future injury rate. The study suggested that such an algorithm could become a promising enforcement tool if used to single out the most hazardous coal mines for extra regulatory scrutiny.
Although reliant upon the same data sources and also intended to improve mine safety and health, the targeting algorithms in this contract research will serve a different purpose. Their intended function is to glean new information on equipment-related hazards that would enable NIOSH to develop more focused and targeted technological interventions. Statistical targeting will be used to help NIOSH define discrete subtypes of equipment-related injuries, understand the prevalence and distribution of these subtypes currently and over time, and determine whether particular equipment-related violations predict equipment-related injuries. This information will provide NIOSH with a more nuanced picture of the overall landscape of equipment-related injuries and some of their underlying causes, which it can then use to design customized safety interventions.
The targeting algorithms, developed by Stanford University, will use several complementary data-mining techniques to glean new insights into the prevalence and distribution of equipment-related accidents, their evolution over time, and whether certain equipment-related violations are predictive of subsequent injuries. It is hoped that NIOSH personnel will be able to use this information to prioritize and develop new technological innovations.
- Characteristics of the Top Five Most Frequent Injuries in United States Mining Operations, 2003-2007
- Designing a Pilot Program for Strategic Mine Safety and Health Improvements through the Use of Surveillance Data to Guide Targeted Inspection Activities
- Ergonomic and Statistical Assessment of Safety in Deep-Cut Mines
- Fall From Equipment Injuries in U.S. Mining: Identification of Specific Research Areas for Future Investigation
- If the Technology Fits: an Evaluation of Mobile Proximity Detection Systems in Underground Coal Mines
- Machine Injury Prediction by Simulation Using Human Models
- Machine-Related Injuries in the US Mining Industry and Priorities for Safety Research
- Mining Fact Sheets
- Mining Roof Bolting Machine Safety: A Study of the Drill Boom Vertical Velocity
- Verification and Validation of Roof Bolter Simulation Models for Studying Events Between a Machine and its Operator