BJA Scale Accuracy Verification
Standards & Teaching Reference

1. Purpose

A laser scanner produces measurements that will be relied on in court. This document states the tolerances applied when verifying that a scanner is measuring correctly, where those tolerances come from, and what the governing standards do and do not require. It exists so that any figure appearing in a verification report can be traced to a published source or to a stated engineering basis.

2. Terminology

Three terms appear throughout this reference and in the reports produced by the verification tool. They name three different things and are not interchangeable. Each resolves to one tolerance.

The assembled multi-scan project is held to ±¼ in per DOJ NCJ 243598. That figure describes the finished product rather than any single measurement, and is reported in the registration report, not by this tool.

The scale bar check is the direct descendant of the accuracy confirmation shot. The arithmetic is identical — vector subtraction and Pythagoras in three dimensions — but it is performed on scan data against a certified artifact rather than in the field with laser and encoder, and it is held to the tighter per-scan tolerance. Where older course material or a legacy report says only “accuracy confirmation,” read it in context; in new material use whichever of the three terms above is meant.

3. The requirement: one quarter inch

The governing federal guidance states a numeric tolerance for crime scene measurement:

“Measurements should be accurate to within ¼".” Crime Scene Investigation: A Guide for Law Enforcement, U.S. Department of Justice, Office of Justice Programs, NCJ 243598, September 2013 — Sketching. The sentence appears twice, under “General Considerations” and under “Creating a Projection Sketch.”

This is the 2013 revision, produced by the National Forensic Science Technology Center under Bureau of Justice Assistance award with National Institute of Justice sponsorship through the Forensic Technology Center of Excellence. The quarter-inch sentence does not appear in the 2000 edition (NCJ 178280), which has no dedicated sketching section. Any citation must therefore carry the year or the NCJ number.

The figure is independently echoed on the OSAC Registry:

“Agency policy shall define the level to which measurements must be recorded (e.g., to the nearest ¼")” OSAC 2023-N-0003, Standard for Diagramming Scenes, Registry Version 2.1 (2025), §7.1.1.

Read precisely: the requirement OSAC imposes is that agency policy define a level. The quarter inch is its worked example. It is strong corroboration that ¼" is the recognised working figure, and it is not itself a mandate.

Corroboration from established instructor practice. The forensic mapping course taught by the author at the Institute of Police Technology and Management (IPTM), University of North Florida, has for several years required students to map outdoor scenes using a total station or laser-based mapping equipment and to demonstrate an accuracy of one inch or better. That requirement is part of the graded testing criteria for the course. The quarter-inch figure applied here is four times tighter than that field criterion, and a terrestrial laser scanner meets and exceeds both.

4. The verification chain

Three tolerances operate at different levels, and conflating them is the most common error.

Verify the instrument at an eighth of an inch per scan, and the quarter inch on the assembled project is met with a factor of two in hand.

5. The one-inch field tolerance, and where it comes from

The accuracy confirmation shot has carried a stated acceptance tolerance in BJA course material since at least 2005:

“Acceptable Tolerance is +/- 1 inch or .083 feet or 2.54 cm” A Guide to Using the LTI TruAngle with QuickMap 3D, Bobby Jones and Associates, Inc., © 2003–2014, p. 2, step 28.d.i.

“Confirm Slope Distance (Acceptable Range +/- .083')” Total Station with PocketZone, Bobby Jones and Associates, Inc., © 2005–2008, p. 1, step 22.f. Step 21 sets the check geometry at a 15-foot default.

0.083 ft is one inch. It is an acceptance gate, not an aspiration: the instrument is measured against a steel tape between two known points and does not touch the scene until it passes.

Why one inch is the right gate for laser-and-encoder mapping

Cross-range error is range × tan(angular error). Laser Technology Inc. publishes the MapStar TruAngle encoder at 0.1° staff-mounted and 0.05° tripod-mounted:

A tripod-mounted encoder reaches one inch of cross-range error at roughly 95 ft; staff-mounted, at roughly 48 ft. One inch is what the instrument delivers across an ordinary outdoor scene, which is why the gate sits there.

6. The scale bar check

A certified reference bar is placed within ten feet of the scanner. Both endpoints are identified in the registered point cloud, their coordinates differenced as vectors, and the three-dimensional distance computed:

d = √(ΔX² + ΔY² + ΔZ²)

The result is compared against the bar's certified length, corrected for ambient temperature where required. FARO publishes a ranging error of approximately ±2 mm at this range; two independent endpoints combine to roughly √2 × 2 mm ≈ 2.8 mm, or about 0.11 in.

On a two-foot bar, a passing result reads between 1.99 and 2.01 feet. That is ±0.12 in — inside half the quarter-inch requirement, and at the instrument's own combined endpoint error, so the tolerance is neither slack nor unachievable.

This is the direct descendant of the accuracy confirmation shot taught since 2005. The method is identical — vector subtraction, Pythagoras in three dimensions, verified nine ways as X, Y and Z against X, Y and Z. What has changed is that the reference is now a certified artifact traceable to NIST rather than a steel tape, and the tolerance is eight times tighter.

7. What the standards deliberately do not say

Outside the DOJ guide, the current standards decline to set numbers, and that is a considered position rather than an omission — accuracy requirements are held to depend on what the measurement will be used to prove.

Where the standard leaves the number to the practitioner, a concrete published threshold is what a training programme is supposed to supply. That is the role these tolerances serve.

8. Supporting standards and references

9. Revision history