Most mine-safety IoT research stalls at prototype stage, scoping review finds

A systematic scoping review of Internet of Things research for mine safety has found that the published academic evidence clusters firmly at the prototype and pilot end of the technology readiness scale, with just one reviewed study reaching an operational demonstration and none providing sufficient evidence for the highest readiness levels. The review, published in Discover Internet of Things, assessed IoT, LoRa, LoRaWAN and related wireless architectures for mine-safety applications in studies published between 2021 and 2025, and coded each study against an operational adaptation of the Technology Readiness Level framework informed by the ISO 16290:2013 standard.

Guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews, the authors retained 50 studies for qualitative synthesis. Of these, 40 supplied enough empirical detail — technological configuration, validation setting, test environment, duration or scale, and performance evidence — to be coded for technology readiness, while 10 were kept as contextual qualitative evidence without a readiness assignment. The resulting distribution is striking: five studies at TRL 3, fourteen at TRL 4, seven at TRL 5, thirteen at TRL 6, a single study at TRL 7, and zero at TRL 8 or 9. In other words, the structured evidence synthesis is concentrated at TRL 3 to 6, the band spanning proof-of-concept through integrated prototypes validated in relevant environments.

The lone TRL 7 study stands out because it documented an integrated mine-IoT system with intended users and functions, operational scale, and more than six months of real-mine operation. Nearby cases illustrate how easily maturity claims blur: one study remained a borderline TRL 6/7 case because its evidence consisted of formal instrumented field testing over two days followed by five months of installation with only episodic follow-up measurements, while another stayed at TRL 6 as an integrated real-mine prototype without evidence of sustained routine integrated operation. A mesh-system study was held at TRL 5 because its complete evidence had to be separated from an underground wireless power transfer subsystem and simulation-based protocol results, and a model-based infrastructure study remained at TRL 4 after evaluation in controlled software and virtual-machine environments with no physical mine deployment.

The review is careful to frame this as a finding about the academic literature, not about industry. The authors explicitly state that the absence of TRL 8 and 9 evidence among reviewed studies is not proof that advanced, certified, commercial or routinely operated systems are absent from mines. Grey literature, patents, manufacturer documentation, certification databases, regulatory sources and internal technical reports were all outside the review’s scope and may contain very different evidence. That distinction matters, because mine operators, technology developers and researchers all need to know whether a reported system can move beyond a controlled demonstration into a mine-specific deployment with acceptable coverage, autonomy, maintenance burden, interoperability and safety design.

The corpus itself reflects where academic attention has flowed. The preserved aggregate operating-context distribution reports underground mining as predominant, with 32 of the 40 TRL-coded studies set underground, six in surface or open-pit contexts, one covering both and one not reported. Gas monitoring was the largest coded application group at eleven studies, or 27.5 percent, covering gases, ventilation-related variables and particulate matter. Worker-centered research included wearables at five studies, tracking and localization at three, rescue-oriented systems at three and smart-helmet implementations at two, with these functional subcategories non-mutually exclusive. Annual counts within the TRL-coded sample rose steadily, from five studies in 2021 and three in 2022 to eight in 2023, ten in 2024 and fourteen in 2025, suggesting growing rather than fading academic interest.

Geographically, the largest coded country descriptor was India, associated with sixteen studies or 40 percent of the sample, followed by China with six and not-reported with four. The review stresses that this concentration characterizes the reviewed academic corpus and its reporting patterns, and should not be read as evidence that India accounts for 40 percent of worldwide mine-IoT implementation or industrial deployment. Because deployment sites and institutional affiliations were inconsistently reported across primary studies, geography is treated as a contextual descriptor rather than a primary quantitative axis, and the preserved coding mixes country, region and multi-country descriptors in ways that could not be fully reconstructed.

Technically, the review treats LoRa and LoRaWAN as related but distinct implementation layers, a distinction often flattened in popular coverage. LoRa denotes the physical-layer modulation used for long-range, low-power radio links, including point-to-point and customized topologies, and within the mining corpus it appeared in tunnel-link validation, worker-safety sensing and propagation-oriented experiments. LoRaWAN denotes the higher-layer network architecture connecting end devices through gateways to network and application servers, enabling centralized device and traffic management, with explicit implementations documented in thermal-monitoring and coverage-oriented network designs. Because the preserved study-level extraction did not support a fully traceable aggregate LoRaWAN frequency, the review reports no corpus-wide LoRaWAN count. The corpus was not LoRa-only: wireless sensor networks, ultra-wideband, Bluetooth Low Energy, ZigBee, visible light communication, DASH7, Wi-Fi, 5G and hybrid systems were all retained when they offered extractable mine-safety evidence.

Propagation physics looms large in why mine environments are so demanding. Underground studies report tunnel-, room-and-pillar-, curve-, obstacle- and line-of-sight-dependent radio behavior, where waveguide-like effects, wall materials, working faces and equipment alter path loss and coverage, while open-cast evidence involves more open geometry and different obstruction effects. The review concludes that no universal propagation model or direct underground-to-surface transfer is supported by the corpus; each context requires separate site surveys, path-loss calibration, antenna placement, link-budget margins and node or gateway spacing. Power supply is treated as part of implementation safety rather than merely a battery-life parameter: the corpus reports mains and battery prototypes, energy-aware sampling, multi-hour wearable autonomy, triboelectric self-powering, wind harvesting with supercapacitor buffering, far-field wireless power transfer as a subsystem, and battery-level prediction, but none offers a general solution for all mines.

The review also identifies persistent reporting gaps. Long-term validation, interoperability, cybersecurity, certification, maintenance and economic viability were inconsistently reported across the corpus. Primary studies sometimes report low-cost hardware, autonomy, communication range or energy harvesting, but the authors caution that these indicators do not by themselves establish total cost, maintainability, cybersecurity effectiveness, hazardous-area safety, certification or operational viability. Prototype functionality, safety-oriented design, intrinsic-safety engineering and formal certification of intrinsically safe equipment are described as distinct evidentiary levels: a successful communication or autonomy test does not establish hazardous-area safety or certification, and monitoring explosive atmospheres motivates the application need without establishing that any equipment is certified.

Methodologically, the review is unusually candid about its own limits. No separate protocol was registered; the documented search framework spanned six primary databases, and a June 2026 retrospective reconstruction documented database-specific yields totaling 394 reconstructed record instances, flowing through 282 records after deduplication and 150 full-text assessments to the 50 retained studies. However, the complete record-level provenance linking reconstructed exports to screening and final inclusion could not be recovered, so these figures are presented as a documented aggregate flow rather than a reproduced screening history. Screening and extraction were led by one reviewer with co-author adjudication, no independent duplicate screening was performed, and a historical 12-study reliability exercise using an unweighted Cohen’s kappa on a small subset is not treated as a reliability estimate for the final classifications. Within those boundaries, the review’s contribution is a structured, criteria-based readiness map showing a heterogeneous academic transition from feasibility toward documented operation — and a research agenda calling for multi-site, long-duration deployments that report users, scale, failures, maintenance, cybersecurity controls, safety qualification and economic indicators needed to move credibly beyond TRL 6.

Subject of Research: Technology readiness of IoT, LoRa and LoRaWAN systems for mine safety in a systematic scoping review of 2021–2025 studies

Article Title: Technology readiness of internet of things systems for mine safety using LoRa technology and LoRaWAN networks alongside related wireless architectures in a systematic scoping review

Article References: Acuña Benites, R. A., & Rosales Huamani, J. A. (2026). Technology readiness of internet of things systems for mine safety using LoRa technology and LoRaWAN networks alongside related wireless architectures in a systematic scoping review. Discover Internet of Things, 6(1), Article 193. https://doi.org/10.1007/s43926-026-00492-7

Image Credits: AI Generated

DOI: 10.1007/s43926-026-00492-7

Keywords: Internet of Things, mine safety, LoRa, LoRaWAN, technology readiness level, systematic scoping review, underground mining, wireless sensor networks, gas monitoring, worker safety, PRISMA-ScR, intrinsically safe design