Is it possible to generate accurate grid lines on the ground for long distances... anyone know?

In my line of work, we roll a GPR cart on the ground in straight lines on both the X and Y axes at 1ft intervals to scan the ground and create a 3D cube from these 2D cross-sections. We usually have to manually put string lines on the ground as a guide to keep our lines straight. I’m curious if AR grid lines on the ground could help make things easier and faster. But I have no idea how accurate those lines would be or how far they could go without drifting. Does anyone think this is possible with current hardware? Also, how much would it cost to hire a developer for something like this? It seems like it could be a simple program.

I was thinking about how you could integrate VI-SLAM with GNSS to attempt it, but honestly, 50m isn’t practical with any accuracy.

What do you mean by VI-SLAM and GNSS? I’m not familiar with those terms.

VI-SLAM stands for Visual-Inertial Simultaneous Localization and Mapping, and GNSS is Global Navigation Satellite System. They’re both used for positioning, but the accuracy can drop over longer distances.

What if something like Dot3D’s Apriltag QR codes system was used? I heard that QR codes could be placed around the job site to help the AR headset scan and orient itself in 3D space.

That sounds interesting. Do you think that would help with accuracy? I’m curious if anyone has tried it.

Yeah, I think it could potentially help with orientation. The AR system could reference the QR codes to maintain its position more accurately.

Have you looked into using LIDAR? I wonder if that could provide more precision over longer distances.

LIDAR sounds like a good option. How does it compare in terms of cost and complexity?

It can be more costly, but it’s usually pretty accurate. You’d need to weigh the benefits against the expenses.

What do you think would be the main challenges in developing something like this? I’m guessing keeping everything aligned is tricky.

Definitely. Keeping everything aligned over long distances without drift would be a major challenge. I’m wondering how much calibration would be needed.

Yeah, I could see that being a headache. It might take a lot of trial and error to get it right.

I’m really interested in this idea. If you find a developer, I’d love to know how much it costs to get this started.

For sure. I’ll keep everyone updated if I find any developers who can help with this. It seems like it could be a valuable tool.