MSHA Compliance Software: Proving a Pattern of Compliance Across Forty Mine IDs

Why an inspection turns into a records hunt

The inspector arrives at Plant 3 at 07:40 unannounced, as they are entitled to do. Within the first hour they ask for workplace examination records covering the last several months, training records for the four people currently on the crusher, and evidence that the guarding condition cited in March was corrected and stayed corrected. The plant manager knows the exams were done, because he watched the foreman do them. The record is a bound book in the foreman's truck, and the foreman is on days off. The training certificates are in a filing cabinet at corporate, 200 miles away, except for two contractors nobody can place. By 10:00 a compliance problem that did not exist at 07:40 exists, because the underlying work was done and the evidence was not producible.

This is the shape of MSHA exposure for a group. It is rarely that safety work is not happening. It is that the evidence lives at site level, in formats that cannot be aggregated, while enforcement lands at corporate level and is assessed across your enforcement history. Section 104 citations carry abatement deadlines. Repeat and unwarrantable failure findings escalate. Pattern of Violations screening looks at your record over time rather than at a single day. A corporate safety director cannot manage that from binders.

The regulations are specific and they are not generic safety practice. Under 30 CFR Parts 56 and 57, working places must be examined at least once each shift, with adverse conditions and the corrective action recorded, and those records retained. Part 46 and Part 48 set different training regimes with different hour requirements depending on your commodity and whether the mine is surface or underground. Part 50 requires accident, injury and illness reporting on its own forms and timelines, including immediate notification of certain accidents. The details do change and they differ by operation, so confirm the current requirements for your mine IDs with counsel or an MSHA specialist rather than with any software vendor, including us.

Problem 1: the workplace examination record is a book in a truck

The once per shift examination is the most frequently produced document in your entire compliance file, and at most operations it is handwritten. Handwriting is not the problem. The problems are that a paper book cannot be searched, cannot be aggregated across sites, cannot prove that a recorded adverse condition was actually corrected, and cannot demonstrate to anyone at corporate that exams are being done consistently rather than being signed in batches on Friday.

Off the shelf EHS platforms will give you an inspection module, and both VelocityEHS and Intelex have capable ones. The friction is that those modules are shaped around OSHA style periodic inspections and generic audit templates, so the mining specifics get configured in as custom forms: the per shift cadence tied to working places rather than to a site, the requirement that the record names the condition and the corrective action, the retention rule, and the fact that the unit of everything in your world is a mine ID rather than a facility. Configuration gets you most of the way and then leaves the aggregation and the reporting shape as manual work.

What a custom build must include is a mobile exam capture that works with no signal in a pit, that defines working places as data so the exam covers the areas actually worked that shift, that turns every recorded adverse condition into a tracked corrective action with an owner and a due date, and that produces the record in the form your operations expect to hand an inspector. Photographs attached to conditions are worth more than any narrative field. Exam completion by site, shift and examiner then becomes a number rather than an assurance.

Problem 2: training compliance is date arithmetic nobody wants to do by hand

Part 46 and Part 48 impose different requirements, and which applies depends on your commodity and whether the mine is surface or underground. New miner training, newly employed experienced miner training, task training when a miner moves to a new job, site specific hazard awareness for contractors and visitors, and annual refresher training all carry their own hours and their own clocks. A single person can be current at one mine ID and not another. Contractors compound it, because a contractor crew arriving Monday needs verifiable training status before anyone signs them onto site.

The build has to model training as competency records with dates and scope, evaluate currency per person per mine ID per task, and block or flag site entry when someone is out of currency. Reminders at 60 and 30 days let a trainer schedule refreshers instead of discovering lapses when an inspector reads a certificate. The training plan itself belongs in the same system with its versions, so you can show which plan was in effect when a given miner was trained. This is exactly the date logic that is miserable in spreadsheets and trivial once modelled properly.

Problem 3: citations run on clocks that start without you

A citation arrives with a condition, a standard cited, a termination due date and consequences that compound. There is a route to a safety and health conference, and there is a strictly limited window to contest a proposed penalty before the Federal Mine Safety and Health Review Commission. Those clocks start at the site, and the decisions get made at corporate, and the information usually travels by photograph of a piece of paper sent to a phone.

The build must treat each citation as a case with the standard cited, the mine ID, the issuing inspector, the abatement action taken, the evidence of abatement, the termination, and every date in the sequence, including the ones your legal team cares about, and it should escalate automatically as a due date approaches. Attach the corrective action to the same record, so abatement evidence is a photograph and a signed action rather than an email thread. The aggregate view then shows which standards are cited most often, which sites repeat the same finding, and where one corporate fix removes a citation being treated forty times.

Problem 4: Part 50 reporting is a deadline problem more than a paperwork problem

Reporting accidents, injuries and illnesses on the required forms is not conceptually hard. What goes wrong is timing and consistency: certain accidents require immediate notification, injury reports are due within a defined period, and quarterly employment and production data has to be filed per mine ID. When those obligations fall to whoever is available at the site, filings go late or inconsistent, and inconsistency is what shows up when your record is examined as a whole.

Build the workflow with the clocks visible: an incident recorded at site triggers a classification step, the classification determines which notifications and forms apply, and the deadline sits on somebody named with escalation if it is not met. Keep the incident investigation on the same record, because the corrective actions it produces are what actually reduce recurrence.

Problem 5: corporate cannot see the pattern until MSHA does

The uncomfortable reality for a group safety director is that the agency has a consolidated view of your enforcement history and you often do not. Violation counts, significant and substantial findings, injury rates and inspection outcomes per mine ID are all part of the record that shapes enforcement posture, and assembling your own version of that picture from site binders takes weeks.

This is the strongest reason to build. A corporate dashboard by mine ID showing exam completion, training currency, open citations by age, repeat standards, incident rates and overdue corrective actions turns an annual scramble into a monthly review. Where machine assistance helps is narrow and honest: clustering the free text of recorded adverse conditions and citation narratives to surface the same underlying hazard described five different ways across five sites. That is a real finding no human spots across thousands of records.

What this costs and how long it takes

Across the 2,000 plus projects Digital Heroes has delivered, this is the honest shape. A first release covering mobile workplace examinations with offline capture and corrective action tracking, training competency by person and mine ID with expiry alerting, and citation case management with abatement clocks, runs $60,000 to $130,000 and ships in 10 to 16 weeks. A full platform adding Part 50 incident workflow, contractor onboarding and verification, equipment defect and pre operational check records, document and training plan control, and corporate enforcement analytics runs $150,000 to $380,000 over 6 to 12 months.

What drives the price up specifically here: the number of mine IDs and how differently each site currently works, because standardising an examination form across twelve plants is a change management exercise before it is a build. Contractor management, which is its own product if contractors rotate often. Integration with an HR (Human Resources) or payroll system so that the roster of who works where is not maintained twice. State requirements layered on federal ones. And offline capability across sites with genuinely poor coverage, which is most of them.

What keeps it down: start with workplace examinations at every site and nothing else. It is the highest volume record, the most frequently requested, and the fastest way to get supervisors using the system daily.

Build versus buy, and when buying is the right call

Buy if you run one or two mine IDs with a small crew. A disciplined paper system with a good filing habit satisfies the requirements, and a custom build would be an expensive way to solve a problem that a binder and a calendar reminder handle. Predictive Solutions is a reasonable purchase if your goal is specifically observation based leading indicator data, which it does well, though it is not a compliance record system and should not be bought as one.

Build when two or more of these are true. You hold more than about eight mine IDs and cannot answer a corporate level compliance question without emailing sites. Your workplace exam records are paper and you have already had an inspection where producing them was the problem rather than the underlying condition. You have been near, or in, elevated enforcement status and need to demonstrate a sustained pattern rather than a snapshot. Your contractors rotate and verifying their training at the gate is manual. Or your current EHS suite is an OSHA shaped product that your team has spent two years bending into mining forms while still keeping a shadow spreadsheet, which is the situation we most often find.

How to choose a developer for MSHA compliance software

Ask them whether they know what a mine ID is before you explain it. The unit of aggregation in this domain is the mine ID, not the facility or the business unit, and a developer who has not built for MSHA will model your group like an OSHA employer and produce reports that do not match the way you are actually assessed.

Ask how the exam app behaves with no signal for a whole shift, and how it prevents a week of exams being entered in one sitting on Friday. The second matters because a record that looks batch entered undermines the whole file.

Ask how they will handle regulatory change. Training hour requirements, forms and thresholds are set by regulation and they move. Rules and forms belong in configuration with effective dates, not in code, so an update is a data change rather than a release.

Ask who owns the code and the data, and get it in writing before kickoff. You should own the repository, the cloud accounts and the right to hire anyone else to continue the work. At Digital Heroes the code is yours from the first commit. Compliance records have long retention obligations and may be needed in a contested proceeding years later, so being able to export everything in a usable form is a requirement rather than a preference.