Request for Guidance on AR-based Indoor Navigation

Hello everyone!

I’m a student currently working on my graduation project focused on indoor navigation using augmented reality. I’m reaching out to see if anyone might have resources, tutorials, or guidance that could help me along the way. If you’ve worked on something similar or know of helpful materials, I’d be grateful for any recommendations!

Thanks so much in advance for any help you can offer!

Best,

You might want to post a few more details about what you want to achieve. I haven’t used AR in a long time, the last time I used it was just for some basic camera base AR. There are already a few documents like AR Overview or AR Foundation to look over.

What kind of equipment do you plan on using (vision pro, iPhone)? It sounds like you might be making a robot that can navigate around a room. In this case a iPhone with spatial camera would probably be good enough. Then just add a navmesh to the generated terrain mesh. Normally you can simulate a lot of this without any equipment e.g. just have a 3d room model.

I’m working on creating an Android app for AR-based indoor navigation. I’ve watched some tutorials, but I have questions about handling different 3D maps. When I bring in a 3D map of a new place, I’d like the app to automatically process the map, identify walkable paths, and set up the navmesh or pathfinding elements without needing to manually adjust each map.

I’m currently planning to use Unity with a 3D model of the environment and possibly rely on navmesh to guide navigation. Is there a way to automate the setup process for new 3D maps so it works seamlessly across different locations?

Regarding equipment, I’m targeting Android devices—would you recommend any specific Unity or AR Foundation techniques for this?

I’m only a novice with AR and I haven’t used it in a long time. If your using Android and AR Foundation then I’m guessing your already using ARCore Extensions.

I thought you might be trying to use spatial mapping to generate a mesh of the room and then dynamically generate a runtime navmesh on that. This would have required devices with hardware support (Hololens, Vison Pro, some IPhones). It would also have issues where you would need to walk around the entire room to get a decent scan.

Sounds like you already have static meshes of your room. I wouldn’t worry about automating the navmesh for that. It should only take a few minutes to generate a new navmesh. I’ve also found that the navmesh isn’t that great at generating so it might take some tweaking to get a good navmesh.

Not sure what your using to generate a 3d mesh of your room. I’m guessing it’s not just manually placing 3d objects as it will unlikely line up with the AR without a lot of work. I guess your generating the mesh using multiple images or from some hardware.

One issue I had with Android in the past was that it couldn’t map very well with the world. When I was trying ARCore I had to add printout markers for it to map to 3d space. I also couldn’t process a full room, just something like an area on a desktop. I’m guessing it’s improved by now but I think you’ll find a new IPhone will be easier to use for AR.