What TR measurement means
TR measurement is a safety observation round on a construction site. The site is walked through and every observation is marked either correct or incorrect in one of seven categories. The marks produce a single number, the TR index, showing what share of the observations was in order.
The abbreviation TR comes from the Finnish word talonrakennus, building construction. Experts at the Finnish Institute of Occupational Health developed the method in the 1990s at the request of the occupational safety authorities, together with construction companies and safety inspectors. TR is a registered trademark, and the Occupational Safety and Health Administration maintains the official description of it. The equivalent for civil engineering sites is the MVR index.
The idea is to turn safety into something countable. Before the method, a site walk produced a list of defects that could not be compared to last week or to the site next door. When every observation is marked either correct or incorrect, the same walk produces a number you can follow week to week.
This guide covers what happens in practice: what the categories contain, how the index is calculated, what counts as an observation, and where measurements typically go wrong. The employer is always responsible for meeting site safety obligations, and this guide does not replace occupational safety instructions.
What it is based on
A Finnish construction site must carry out a maintenance inspection at least once a week. The obligation comes from the government decree on the safety of construction work (205/2009), and the inspection covers among other things site order, fall protection, lighting, electrical installations, lifting equipment, scaffolding and access routes. Records of the inspections must also be kept.
TR measurement is not itself a statutory method but the established way of carrying out that weekly inspection. Its popularity rests on two things. The walk happens anyway, so making marks during it costs almost nothing. And because the method is the same everywhere, results are comparable between sites and companies.
In practice many clients and main contractors also require TR measurement in their contracts and ask for the index regularly. It is often a contractual obligation even though the method itself is not named in law.
The seven categories
The categories are the same on every TR form. Below is what each contains and what actually decides the direction of the mark.
1. Working. You look at people, not structures. Is protective equipment in use and worn correctly, is the work done safely, is unnecessary risk being taken. Each worker is a separate observation.
2. Scaffolding, walkways and ladders. Condition and anchoring, guardrails in place, proper access, ladders of the right type and used correctly. Working from a leaning ladder is one of the most common incorrect marks.
3. Machines and equipment. Condition, guards in place, safe positioning. This covers both plant and hand tools.
4. Fall protection. Guardrails, openings covered and marked, harnesses where that is the solution. This is the category where the most serious defects usually are.
5. Electricity and lighting. Distribution boards, cables and their routing, general and task lighting. An extension lead across the floor belongs here, not under housekeeping.
6. Housekeeping and waste. Clear access routes, material storage, waste sorting and removal. In practice this correlates most strongly with the other categories: a messy site usually has other problems too.
7. Dust. Dust spreading, extraction in use, standard of cleaning. This is the newest of the categories: it was added to the method in 2010. It matters most in renovation work and during interior finishing.
Go through the categories in the same order every time. That is not a formality but a way to make sure no category quietly gets less attention and pushes the result upwards.
How the TR index is calculated
The formula is simple:
correct / (correct + incorrect) x 100 = TR index
If a round produced 90 correct marks and 10 incorrect ones, the index is 90 percent. It is the share of correct observations out of all observations.
Two things about the formula are worth understanding, because they explain its limits.
The index knows nothing about severity. One missing guardrail and one badly stacked bundle of boards are both a single incorrect mark, even though one can kill and the other can bruise. That is why incorrect marks are always written out: what was wrong, where, who fixes it and by when. A percentage on its own steers nothing.
The index depends on the number of observations. On a small site with 30 observations, one incorrect mark drops the index by more than three percentage points. On a site with a hundred, the same mark costs one point. So do not compare indexes between sites of different sizes; compare the same site from week to week.
The index runs from zero to a hundred percent, and readings above 90 percent are considered a good level. That figure is now a minimum criterion in many client contracts, meaning a number you are not allowed to fall below. Some companies set their own target higher, at 95 percent for example.
What counts as an observation
This is where TR measurements differ most from one another, and if it is not agreed, the numbers are not comparable even within one company.
The rule is: a mark is made per observation, not per object. The same scaffold can produce several marks, because each separately observed thing is recorded on its own. If the scaffold has a proper guardrail but no toe board, it yields both a correct and an incorrect mark.
Likewise each worker is a separate observation under Working. On a site with ten people that alone produces ten marks, even when everyone is doing the job correctly.
The practical consequence: a round produces more marks than objects, and on an ordinary site the count lands somewhere around a hundred. If a round produces 15 marks, it was done too superficially, and the index mostly reflects how fast someone walked.
How the round is done
Who. The round is done together, not alone. It typically includes the site manager or a supervisor and a workers' representative, such as the elected safety representative. Doing it together matters for two reasons: two pairs of eyes see more, and the observations are agreed on the spot rather than afterwards by email.
When. The same weekday and roughly the same hour every week. Friday afternoon and Monday morning give very different pictures of a site, and if the timing moves, part of the change is just that.
Route. The same route every time, the whole site. If the round regularly skips the same basement, that basement never shows up in the numbers.
Recording. Marks are made during the round, not from memory afterwards. On paper that means tally marks, on a phone a tap. Either works as long as it happens in the moment.
Immediate danger is fixed immediately. If the round finds something that could cause a serious accident, it is fixed or cordoned off on the spot. The observation does not wait for the report. This is the most important rule of the whole round, and the one most easily forgotten when attention goes to filling in the form.
The result is told to the crew. The index and the items to fix are gone through with the site. If the result stays in the supervisor's folder, the measurement changes nothing and becomes a chore everyone recognises as pointless.