Montréal’s Réseau express métropolitain, or REM, is a 67 km (41.6 mi) automated light metro network across Greater Montréal, with 26 stations planned across the system. Built across tunnels, elevated guideways, bridges, and major station environments, it’s one of the largest transit infrastructure programs underway in North America and one of the largest automated networks of its kind.
For NouvLR, the engineering and construction general partnership building the REM, documenting a project like this meant working across a wide range of environments while keeping project data reliable enough to support design, construction, verification, and delivery.
How they did this was the focus of a LinkedIn Live hosted by NavVis, featuring Alicia Llorens from NavVis, Jacobo Trigo from NouvLR, and Simon Gingras-Gagnon from Cansel. Together, they showed audiences how NouvLR brought NavVis into a broader surveying and reality capture workflow to capture difficult spaces more efficiently while maintaining confidence in the data.
The scale of the REM helps explain the challenge: 67 km (≈41.6 mi) of trackworks, 16.2 km (≈10.1 mi) of elevated guideway, 26 stations, 25 bridges, 3 MSF yards, a 4 km (≈2.5 mi) TBM tunnel under Montréal airport, and another 5 km (≈3.1 mi) of tunnel renovation. It also includes 11 bus terminus connections, 3 metro connections, 14 park-and-ride facilities, 2 train connections, and 1 airport connection.
And of the 26 stations, 7 are elevated, 13 are ground-level, and 5 are underground.
In particular, on the REM project, reality capture sits inside a much broader project discipline. Jacobo Trigo, who leads work related to construction support and surveying at NouvLR, outlined a team responsible for helping bridge design and construction and for checking that what was built matched what was designed.
At peak, his team included around 125 people, and to support their work, NouvLR installed more than 5,000 control points and thousands of levelling loops across underground, elevated, and ground-level conditions.
The foundation they laid shaped how NavVis was to be used on the project. The capture workflow was tied into a dense control network, a layered QA process, and a broader effort to maintain a current, usable picture of the project as it evolved.
NouvLR described it all as a living 3D as-built model built from point clouds, conventional survey, videogrammetry, and design-intent models. And from there, the task became building and maintaining a record that teams could use for coordination and verification over time.
By the time NavVis came in, NouvLR already had a mature capture environment in place through its collaboration with Cansel. The team was already working with tools including GEDO Trimble systems and DJI M350 + L2.
NavVis came in, however, to fix a clear gap in their laser scanning workflow: Confined TBM corridors, where working windows were limited and accuracy still mattered. For these areas especially, the team needed a faster way to collect data, reduce revisits, and keep additional field work under control.
So NouvLR tested NavVis MLX against the standards already in place on the project. The team compared NavVis MLX data with datasets from TX8 (2022-09-07) and GEDO Scan (2025-08-14).
The visible measured differences on those slides were in the low millimeter range, including 0.002 m (0.08 in), 0.008 m (0.31 in), 0.009 m (0.35 in), and 0.016 m (0.63 in) in one comparison, and 0.004 m (0.16 in), 0.007 m (0.28 in), and 0.011 m (0.43 in) in another.
All in all, clear evidence that NavVis could not only improve speed in the field, but fit into the project’s existing quality framework.
| Comparison | Visible measured differences |
|---|---|
| TX8 (2022-09-07) vs. NavVis MLX (2025-11-25) | 0.002 m (0.08 in), 0.008 m (0.31 in), 0.009 m (0.35 in), 0.016 m (0.63 in) |
| GEDO Scan (2025-08-14) vs. NavVis MLX (2025-11-25) | 0.004 m (0.16 in), 0.007 m (0.28 in), 0.011 m (0.43 in) |
The rollout moved quickly. With support from Cansel, the timeline ran from an initial discussion on November 10 to completed scans by February 18. In between came the first demo, a scan window on the tracks, delivery of a registered and cleaned point cloud, data comparison, on-site training, and then the first scans by NouvLR’s own team.
The workflow was integrated carefully from the start.
Zones were segmented in advance, devices were selected according to site conditions, and collection was confirmed on-site before teams left the area. NouvLR also pointed out how helpful overnight processing in NavVis IVION was as it helped shorten the path from field capture to quality validation dramatically.
In about 2 months, with the help of NavVis and Cansel, NouvLR had scanned, processed, cleaned, and georeferenced approximately 50,000 m² (≈538,200 sq ft).
That included 7 full stations, 1 MSF, and 3 km (≈1.9 mi) of tunnel corridor, along with parking and mechanical rooms, all captured under night shifts only and within very limited scanning windows.
The largest individual figures shown by Jacobo in the webinar included 12,056 m² (≈129,770 sq ft) for Station McGill and 8,806 m² (≈94,790 sq ft) for Station Édouard-Montpetit. Other locations included Station Technoparc, Station de Correspondance A40, Station SADB, Corridor TBM, Station Montpelier, MSF SADB, Station Canora, and Station Mont-Royal.
| Item | Figure |
|---|---|
| Total area scanned, processed, cleaned, and georeferenced | ≈50,000 m² (≈538,200 sq ft) |
| Detailed total shown | 47,060 m² (≈506,550 sq ft) |
| Full stations covered | 7 |
| MSF covered | 1 |
| Tunnel corridor covered | 3 km (≈1.9 mi) |
| Working conditions | Night shifts only, very limited scanning windows |
| Location | Area |
|---|---|
| Station McGill | 12,056 m² (≈129,770 sq ft) |
| Station Édouard-Montpetit | 8,806 m² (≈94,790 sq ft) |
| Station Technoparc | 3,750 m² (≈40,365 sq ft) |
| Station de Correspondance A40 | 3,582 m² (≈38,556 sq ft) |
| Station SADB | 3,410 m² (≈36,705 sq ft) |
| Corridor TBM | 3,040 m² (≈32,722 sq ft) |
| Station Montpelier | 2,779 m² (≈29,913 sq ft) |
| MSF SADB | 2,743 m² (≈29,526 sq ft) |
| Station Canora | 2,156 m² (≈23,207 sq ft) |
| Station Mont-Royal | 2,107 m² (≈22,680 sq ft) |
Of note, Édouard-Montpetit is a useful example of the extreme project conditions. Described by Jacobo and Simon both during the session as the deepest station in Canada and possibly the second deepest in North America, it brought together vertical complexity, multiple levels, limited access, and difficult field logistics.
The REM gives a good sense of what reality capture has to do on a project of this size. It has to fit into the wider structure of surveying, quality validation, and project delivery. It also has to produce data that teams can use while work is still moving.
That is where NavVis found its place in NouvLR’s workflow. The project already had a strong control network and a clear quality process behind it, and NavVis helped the team work much more efficiently in difficult spaces while staying within that structure.
In the Q&A, Jacobo said it was difficult to quantify the savings exactly, but compared with more traditional TLS-based methods, they were not less than 4x or 5x.
That matters on a project like the REM because time in the field is only part of the equation. The data also has to remain reliable enough to support the next decisions, the next checks, and the next stage of work.
That is what made NavVis MLX, together with Cansel’s support, so valuable to the project.
Ready to see how NavVis can transform your own workflows?