Picture the last topographic plan your team drew for an urban block. Somewhere in the middle of it, a person scrolled through a point cloud, located every lamp post, drain, sign and manhole cover, and dropped a block on each one. Skilled people. Repetitive work.
That’s the job some drafting teams are now handing to software. Here’s why the economics are starting to make sense.
The staffing squeeze is real
A 2025 survey of more than 600 design and construction professionals in the US, Canada, the UK and France, run by project-information firm Newforma, found that 41% of firms had seen project delays because of labour shortages. Geospatial work has its own version of the problem. Speaking at Geo Week 2026, analyst Shawana Johnson said the geospatial workforce is roughly 10 to 15 years behind where employers need it to be.
When you can’t hire your way out of a backlog, you look for the repetitive part of the work and ask whether a machine can do the first pass.
What the first pass looks like
Point-object detection is a good candidate because the task is narrow and the objects keep recurring. The model proposes where the lamp posts, drains and hydrants are. A drafter confirms, fixes the odd miss and converts the detections into the symbols the client’s standard calls for.
The human doesn’t disappear. The work shifts from finding to checking, which is faster for the same reason proofreading is faster than writing.
One tool built for it
GIS J.O.T’s VisionSpot AI is built for exactly this step. It reads point clouds in E57, LAS and LAZ, detects streetlights, road signs, utility poles, manhole covers, storm drains, trash cans and fire hydrants, and runs as a plugin inside AutoCAD and Civil 3D. The company’s own engineers use it on live projects and report time savings from 15%, depending on project size. That’s a vendor-reported figure, so treat it as a claim to verify on your own data.
Where the hours actually go
On a point-object job the time splits three ways: finding the object, placing it precisely, and converting it into the client’s symbol. Detection takes over the first. Calibration to a drafting standard takes over most of the third. Precision placement still deserves a person’s eye, because a block that sits a metre off costs more than one that’s missing.
There’s a side benefit too. A tool configured once applies the same standard to every project, so symbols don’t drift between drafters. That’s a quiet cost saver on handover, when someone else has to open the file.
Four questions to ask before adopting anything like it
- Does it read what your scanners produce? File formats and size limits decide whether a pilot even starts.
- Does it speak your standard? Detections are only useful if they arrive as your symbols, on your layers.
- Who reviews the results? Look for a workflow where a person accepts or rejects detections before they reach the drawing.
- Can you test it on your own file? A demo on someone else’s clean data tells you little.
A pilot that costs nothing
On the last point, VisionSpot’s developers will process a sample file you send, configured to your drafting standards, free of charge. One project’s worth of data is enough to see whether the hours really move.
Labour will stay tight. The teams that cope best will be the ones that decided early which parts of the job people should still do.
Hi, I’m Tony — a passionate blogger with over 3 years of experience in writing informative and accurate content. I specialize in sharing practical insights on sizes, measurements, and spatial guides to help readers make confident decisions. Through DimensionsPoint.com, I aim to simplify complex data into easy-to-understand content that’s reliable, useful, and SEO-friendly.
When I’m not writing, I’m researching the latest trends in measurement standards and user needs to keep my content relevant and up to date.