FROID NR-1/ISO-45003
Instrument anchored in ISO 45003:2021 — the backing from an international institution that regulators require of an assessment tool
Reference regime: Brazilian rule NR-1 as amended by Ministerial Order MTE no. 1,419/2024, in force since 26 May 2026
| Client | |
|---|---|
| Company registration no. | |
| Site assessed | |
| Workers at the site | |
| Date |
For Brazilian operations: the advisory window closes on 24 August 2026
The labour ministry's standards office answered, on the entry into force, that in the following 90 days the Labour Inspectorate tends to prioritise guidance, instruction and notice — “without prejudice to administrative measures in applicable cases”. It is not an exemption; it is a declared priority, and it has an end date.
| 27/08/2024 | publication of Ministerial Order MTE no. 1,419, which rewrote chapter 1.5 of the rule |
|---|---|
| 26/05/2026 | full effect, after the postponement in Ministerial Order MTE no. 765/2025 |
| 24/08/2026 | end of the 90 days of advisory double visits |
| from then on | a citation under art. 201 of the labour code, with a fine of R$ 2,396.35 to R$ 6,708.08 under NR-28, plus notice and a requirement to comply |
The fine is the smaller part. What weighs is what it documents: from the citation onwards there is an official record that the company was notified and did not act — and it is that record that reappears in the employment claim brought by the employee who fell ill, at the stage where the judge asks whether the company knew, whether it should have known, and what it did about it.
Outside Brazil the dates and the fine do not apply, and the rest of this proposal does: what transfers is the instrument, the floors, the effectiveness engine and the evidence trail. The deadline that binds you is your own jurisdiction's, and your counsel states it.
Starting now means starting in remediation, and that counts in your favour
The rule does not punish lateness in itself: it demands technical coherence and effective implementation. The inspectorate says so itself — the analysis “will prioritise the coherence between risk assessment, measures adopted and their effective application”, and not the mere formal existence of documents.
In practice, a company that opens the cycle after the deadline and reaches inspection with a dated inventory, an action plan with an owner and a schedule, and evidence that collection took place, is in a different position from one that has nothing — even having started late. What cannot be recovered is time: the baseline only exists from the day it is measured, and without a baseline there is no way to demonstrate effectiveness later.
That is why delay makes the next cycle more expensive, not this one. Whoever measures today can prove improvement a year from now. Whoever measures a year from now will only have something to compare in two.
“And do you prove it worked?”
That is the question inspectors and courts ask — not “did you run the questionnaire”. The rule requires the grading of risk to take into account the effectiveness of the measures already implemented. Measuring once satisfies half the obligation.
A complete psychosocial risk management cycle, and not the administration of a questionnaire. The rule does not end at the survey: it requires the grading of risk to take into account the effectiveness of the measures already implemented (clause 1.5.4.4.5.3) and requires ineffective measures to be corrected (1.5.5.3.2.1). That can only be demonstrated by measuring twice.
Verification of size per site, existence of a worker support channel and of the grading matrix already used in the risk-management programme. Determines whether the route is the full assessment or the preliminary ergonomic assessment.
Timescale: 5 to 6 weeks from signature to delivery of the documents.
Custody of the recorded criteria, the evidence and the dates during implementation of the measures, which is the client's responsibility. Access to the dashboard and to FROID Explains.
Duration: 6 to 12 months.
A second campaign with the same instrument and the same breakdowns; comparison with effect size and confidence interval; update of the inventory and the action plan; entry in the 20-year history (1.5.7.3.3.1).
No steps by band. The price follows headcount continuously — growing from 300 to 301 workers does not change plan and does not reprice the contract. All amounts are in Brazilian reais (R$), the currency in which FROID invoices.
| Component | Amount |
|---|---|
| Platform base, per establishment | R$ 200 / month |
| Band 1 — from 1 to 100 workers | R$ 15.00 / worker / month |
| Band 2 — from 101 to 300 | R$ 12.50 / worker / month |
| Band 3 — from 301 to 1,000 | R$ 9.30 / worker / month |
| Band 4 — above 1,000 | R$ 6.55 / worker / month |
The bands are cumulative, like a tariff: a company with 300 workers pays R$ 15 for the first 100 and R$ 12.50 for the next 200.
| Workers | Monthly | Annual | Per worker / year |
|---|---|---|---|
| 100 | R$ 1,700 | R$ 20,400 | R$ 204.00 |
| 300 | R$ 4,200 | R$ 50,400 | R$ 168.00 |
| 1,000 | R$ 10,710 | R$ 128,520 | R$ 128.52 |
| 3,000 | R$ 23,810 | R$ 285,720 | R$ 95.24 |
| Item | Amount |
|---|---|
| Implementation and baseline — up to 300 workers | R$ (reference: R$ 8,000 to 15,000) |
| Implementation and baseline — above 300 | R$ (reference: R$ 25,000 to 45,000) |
| Proof-of-effectiveness campaign | included in the cycle |
| Cycle term | months |
With realistic participation of 65%, a campaign starts producing a releasable result at around 210 workers on site. That is a participation projection, not a size limit: what the rule requires is a number of answers, not a size of company. A site from 15 workers upwards publishes provided everyone answers — in that range the required sample is already the entire headcount, and a single refusal suspends the inventory. There are two floors: anonymity requires 15 answers, and representativeness requires the sample that speaks for the declared headcount — below 98 people that sample becomes a census, meaning the whole site answering. In that case the correct route is the preliminary ergonomic assessment, mandatory for every organisation with employees and not dependent on any floor.
| Ergonomic assessment with inventory and action plan | R$ / year (reference: R$ 6,000 to 12,000) |
|---|
Observation of the real activity, dialogue with workers and the execution of prevention measures belong to the contracting company. FROID does not conduct field assessment, does not perform an AET and does not implement any measure — neither with its own team nor through third parties.
What FROID delivers is the instrument: the measurement, the record of each piece of evidence with its declared method, and the verification of the effectiveness of the actions the company decides and carries out, in the form the rule requires.
The reason is technical, not commercial. The company is the one that knows the activity, the organisation of work and the concrete conditions — and it is to the company that the rule assigns the decision on measures, their implementation and the signing of the documents. A supplier that proposed and executed its own measures would be measuring the result of its own work.
We say this before the engagement, not after. Selling a campaign to a company that lacks the size for it produces an empty dashboard and a client who is right to complain.
| Worker support channel | A real, accessible and confidential service, with a name and a means of contact. Without it the system does not open the campaign. Asking someone how they are without having anywhere to refer them is worse than not asking. |
|---|---|
| Organisational structure | Sites, departments and headcount per department. |
| Grading criteria | The matrix already used for the other risks in the risk-management programme, so that psychosocial risk is graded coherently. |
| Legal basis and data protection officer | Appointment of the data protection officer and validation of the legal basis for the processing by the client's counsel. |
| What is usually delivered | What FROID delivers |
|---|---|
| One measurement, one report, one file. | Two measurements and the comparison between them — which is what the rule asks for and what sustains the defence. |
| A percentage and a traffic light. | Effect size and confidence interval. A handsome report is not proof; proof has an interval. |
| A conclusion written by whoever provides the service. | A verdict calculated by a criterion fixed before collection. If the result depends on who writes it, it is not a measurement — it is an opinion with a chart. |
| A confidentiality promise in a contract. | Technical impossibility: the database does not allow an individual answer to be queried, by anyone — not by the employer, and not by us. |
| An instrument adapted to each client. | A fixed instrument. Without it, nothing is comparable between one measurement and the next, and the proof of effectiveness ceases to exist. |
| A result that always indicates improvement. | A result that says “no change” when the data does not support improvement. A gain that the statistics will not carry is what an expert takes apart at a hearing. |
The processing is based on the employer's compliance with a legal obligation arising from the occupational health and safety rule — and not on the worker's consent, which in an employment relationship would be vitiated by hierarchy. The worker is informed of the purpose, not asked to agree to it. Under Brazilian law this is LGPD, art. 7, II and art. 11, II, “a”; in another jurisdiction the equivalent basis is stated by the client's counsel and recorded in the engagement.
FROID produces an assessment of working conditions. It does not make diagnoses, does not assess individuals, does not classify a worker into a risk band, is not a validated psychometric instrument and does not replace clinical assessment. Its results may not be used in decisions about hiring, promotion, dismissal or any individual measure.
The signatory declares that they hold powers to represent the client and to bind it under the terms of this proposal.
Proposal valid for 30 days. The actual engagement is governed by the General Terms of Use and the SaaS Licence Agreement in force, accepted electronically in the system with a record of the version and of the exact text presented. This English text is a translation for convenience; in case of divergence, the Portuguese version of this proposal and of the contractual documents prevails. · Full process · Readiness check