The violation itself
The code, the description, whether it was an out-of-service violation, and the inspection it came from.
Free report · Public FMCSA data
We read the same public FMCSA files your safety score is built from, and show which violation moved your score, by how much, and how long you have to ask FMCSA to review it. The report is free. You file the request yourself.
One illustrative carrier, drawn from the shapes that actually occur in the June 2026 files. Every line below is a value computed from the public data — there is nowhere in the report for prose to be invented.
Where the 142 points come from
The intervention threshold is a hard line. Our estimate is not a hard number, so we draw the range we measured around it. When the range crosses the line, we say so instead of picking a side.
This is our estimate from public FMCSA data, not an official FMCSA percentile. The range is the largest error we measured against FMCSA's own published figures, not the average, and it is remeasured every month.
FMCSA counts at most 30 severity points from any single inspection. Past that line the extra violations are already not counted — so removing one of them changes your score by nothing at all.
Everything below comes out of public data you could download yourself. What is hard is not getting the files — it is knowing what the numbers do once they are in there.
The code, the description, whether it was an out-of-service violation, and the inspection it came from.
The marginal contribution of that violation to your BASIC measure, accounting for severity, the out-of-service addition, the age of the inspection, and the 30-point inspection cap.
Including the case where the cap was already reached and taking the violation off would change your measure by nothing at all.
Our estimate of your position against the intervention threshold, shown as a range rather than a single number. This is our estimate from public FMCSA data, not an official FMCSA percentile. FMCSA does not publish percentiles for property carriers at all.
The filing window that applies to this inspection, so you can decide before the question is moot.
A completeness check of your first Request for Data Review, against the documents you upload. One payment, no subscription.
Upload the inspection report, the citation, repair invoices, court records. Every field we read out of them is shown next to the part of the document it came from.
Nothing reaches your draft until you confirm it. A field you did not confirm cannot appear in the text — that is enforced in code, not by review.
Whether your situation matches the official Request for Data Review categories, and which evidence each category expects. If two readings of your documents disagree, we show you both instead of picking one.
Which dates apply to your filing under the rules in force on the day you file, including the DataQs changes scheduled for September 2026.
Assembled only from the fields you confirmed, carrying the signature warning and the 18 U.S.C. 1001 notice that the filing puts on you.
This list is not modesty. Each line is a boundary we designed the product around, and you should hold us to it.
Sometimes the honest result is that you have no grounds to file. That is a finding, not a failure, and it is included in the $79 — we say so and explain which category you fall outside of.
It is also the most common correct answer in this domain. Spending a week assembling a request that gets closed as unfounded costs more than knowing on day one.
You are being asked to trust a number about your own business from someone who emailed you out of the blue. Here is what that number is made of.
RoadsideProof is operated by SOV17 LLC, 3 Moskovyan St., Kentron, Yerevan 0001, Republic of Armenia.
Questions, complaints, or a request to never hear from us again: support@roadsideproof.app. Every email we send carries a one-click unsubscribe, and it works immediately.