An AR/VR product I’d love to develop. What are your thoughts?

As some of you will have spotted in past articles, I’ve spent the last couple of years working with solutions at the intersection of OSS/BSS (data) and AR/VR for visualisations and new ways of working.

A couple of animated GIFs from that work are shown below.

An interactive 3D model with assets identified from OSS / inventory data.

An AR projection of a tower onto our boardroom table (apologies for my shaky capture!). These two images were both scanned using a drone with high-resolution cameras onboard (not that you’d realise from the low-resolutions allowed in animated GIFs as shown).

The cool thing about the 3D model (top image) is that you can zoom right in and take readings (eg the length of one of the concrete plinths) with sub-centimetre accuracy. This can be really useful for remote site visits (eg if the site is really distant, it has OH&S issues, getting site inductions done is painful/expensive, etc). It also allows remote operators to get an immersive experience of what the site conditions are. This can be really useful for when remote workers wish to communicate with workers who are currently on-site.

The downside of capturing these 3D models is that it might cost thousands of dollars per site because of the overhead costs of the drones and needing to have a licenced drone pilot to capture the imagery.

There are certain scenarios where those types of costs aren’t viable, including:

• When conditions on site change. Perhaps a tree falls on a site, a vehicle crashes into cabinets on site, a natural disaster like a tornado strikes, or many other possible events. A worker in head-office urgently needs to get current situational awareness of what’s happening at the site but the telco doesn’t have an available tech (or drone + crew) within hundreds of kilometres of the site
• A telco has thousands of small-cell sites (ie small-cell equipment mounted on power / utility poles) and wishes to create an accurate 3D site survey of every site for roll-out design / planning purposes, but can’t justify the expenses of a survey crew visiting each site (also often ranging into the thousands of dollars)
• There are restrictions on flying drones in certain areas such as CBDs, near airports or rail corridors where protection zones need to be sought from regulators (often introducing high costs and time delays)

This is where the product idea comes in – I think there could be a viable business to create a crowd-sourcing 3D site capture app. Using the app, a company such as a telco or utility (or many others) raises a job to digitise a site (based on geo-coordinates) and offers a image capture bounty. Anyone who visits the site and captures / uploads suitable imagery will be awarded the bounty via the app.

Previously, this type of imagery was really only possible via drone or chopper-captured pics.  But now using NeRF (Neural Radiance Field) technologies, it’s possible to capture 3D models from modern smartphones. This democratises crowd-sourced image capture and potentially gives organisations like telcos much faster situational awareness at a much more viable price-point (albeit at a lower resolution / accuracy / coverage of a drone image set).

If you’d like to see how NeRF technology can achieve this, check out the video below.

What do you think? Is this something you’d like to develop? Is it something you’d like to use (either as a buyer or seller of 3D models)??