Industry News, Latest News, Technology

Reveal: Seeing, is believing

The Reveal platform in action, clearly showing subsurface utilities. Images: Reveal.

Reveal’s research and development of reality capture technology looks set to change the world of utility visualisation and management for the better, and forever. Roads & Infrastructure speaks with Reveal CTO Tim Rastall to learn more.

Tim Rastall, Chief Technical Officer of Reveal, is passionate when it comes to the development of new technology. With experience in design, consultation, gaming and development, he leads Reveal’s efforts in identifying, testing and using nascent technologies to make the underground more knowable.

Now he’s deploying his experience and practical skills on the world of reality capture. 

For the infrastructure and construction sectors, reality capture is a radical new approach to recording highly accurate information on the location and conditions of assets that is typically too costly or time-consuming to obtain.

Rastall says, “Historically, the location of infrastructure was recorded on paper plans by contractors just before the ground was closed, or simply copied over from the engineering plans with no verification at all. Most of the utility plans we encounter are outdated, incomplete or inaccurate in some form.”

“For us, photogrammetry is an enabling technology that solves what we as a business are aiming to solve; that is people being aware of what’s underground.

“Reveal’s focus isn’t on building photogrammetric algorithms, but it’s about leveraging that technology while ensuring that the data capture process is easy and can be made accessible and intuitive to everyone who needs it.”

Example of a photogrammetry scan created from taking multiple still images from a mobile camera.
Example of a photogrammetry scan created from taking multiple still images from a mobile camera.

From its headquarters in New Zealand, Reveal offers a platform for combining and visualising subsurface infrastructure data from sites or projects worldwide. The Reveal platform combines technology, data and field services to help build up the most accurate, comprehensive picture of the underground utilities and risks in a given area.

The company’s technology has already been deployed on infrastructure projects throughout New Zealand and has been acknowledged as a key contributor to productivity gains and cost savings on major infrastructure developments.

“One of the biggest reasons why we’re doing this is that every year, dozens of people die in utility strikes (an incident where a tool or worker damages an underground utility) and thousands more are injured. If we could record information around the position of every asset that has ever been exposed to the naked eye, within 10 to 15 years no more shovels will need to go through another main in the wrong spot, ever again.”

Rastall is leading Reveal’s research into reality capture – the practice of using image sensors to record the true reality of a physical environment using the principles and physics of light.

One familiar technology, LiDAR, uses a rotating laser to scan a surrounding area and determining distances by measuring the time required for the light to reflect back from an object. This is either done using as static LiDAR sensor that rotates about a fixed axis or a handheld/mounted unit that is moved around the area under survey.

Photogrammetry, on the other hand, builds up a realistic, three-dimensional model of an area using still photos captured with inexpensive digital cameras.

“Whenever a technology comes around that’s transcendent and transformative for the good of an industry, where it can result in making things more efficient and can help people, that always gets me excited,” Rastall says.

Reveal has already conducted experiments with reality capture, using photogrammetry and LiDAR separately, and in combination, to rapidly capture as-built info from exposed utilities, chamber interiors and surface models of the road corridor. These experiments have led to the use of both technologies in day-to-day utility surveys as well as Reveal now offering services to verify, refine and package the resultant data collected by its customers.

“What we do as a business is not exclusively reliant on photogrammetry or reality capture, but it’s an increasingly big part of our goal to turn raw data into something that’s meaningful to a contractor, consultant, or utility asset owner. For us, it’s a big paradigm shift in the way information about utilities are recorded,” he says.

When looking further into the realm of reality capture, the Reveal team realised that what previously could only be achieved with high-quality cameras and desktop software, could now be done with the convenience of one’s smartphone.


 

Related stories:

 


“It’s rare that you get the opportunity to collect visual information about the subsurface, as assets spend the majority of their lifespan buried under metres of dirt and concrete. Of course, visual evidence is the gold standard that can remove most of the ambiguity around the location and condition of assets. But no one is going to wait to fill in a road, that might have a lane closed, so we can arrive with survey gear, unless it’s a project deliverable,” Rastall says. “But if the team that is doing the work on site can pull out their phone and spend five minutes scanning the exposed assets, we see that there’s no reason that information can’t be captured.

“The utility locating industry has developed a suite of tools, standards, and practices that can be used to piece together a good estimation of where buried assets are, but there are inherent limitations to these non-invasive locating practices. However, we can eliminate any remaining ambiguity when an asset is exposed by excavation. The opportunity that we saw from smartphone-based photogrammetry is that the value of information about exposed assets in the context of the utility locating industry is enormous.”

Reveal currently uses an application called Pix4D, a combination of Pix4D Catch (phone app) and a cloud-based application called Pix4D Cloud. With a small amount of training and additional equipment, the Reveal team can quickly record high-quality data for both exposed assets and the interior of subsurface structures, such as chambers. 

Rastall admits that there are still some associated challenges. One is accurate georeferencing (putting the scan in the right place on a map), as a smartphone’s GPS can’t provide sufficient accuracy for this purpose and an incorrectly placed scan diminishes its usefulness, particularly when it is otherwise the best source of information available about the assets that were scanned. 

Tim Rastall, Chief Technical Officer – Reveal.
Tim Rastall, Chief Technical Officer – Reveal.

Based upon the trends he’s been observing, Rastall believes the majority of georeferencing challenges will be solved in the next five years.

“We want things like reliable georeferencing to be as automated as possible and there’s a few emerging technologies that we’re watching that might be able to solve that. One is a Visual Positioning Systems, that uses real time imagery from the surroundings to localise the smartphone. Advances in artificial intelligence will start to have an impact as well,” he says.

Environmental conditions can also play a part; if a surface is wet, reflective or dark, it may not be suitable for photogrammetry, but AI is likely to help minimise the effects of these conditions.

“The quality of the image that you’re collecting directly correlates to the quality of the scan,” Rastall says. “But it looks likely that AI will eventually provide the means to detect and repair these kinds of issues and that’s already happening in fact.

“In the next few years, I expect that there’ll be some real shifts in the uptake of this technology. Currently, whenever we do a photogrammetric scan, we include a ruler to confirm that it’s scaled appropriately once processed and to prove that the scan is dimensionally accurate. It’s remarkably precise for the fact that it’s compiling a bunch of photos and building a three-dimensional model from them. It’s almost magical in terms of its ability to do that.”

This high-fidelity information can then be used almost immediately in projects by importing the models into design software packages so engineers can create more efficient and higher quality concept and design drawings. 

Further improvements to reality capture technology will also be spurred by more people using similar platforms. Rastall uses the development of the Pix4D mobile app as an example.

“When it’s in your pocket and can provide real time feedback to ensure your efforts result in a good scan such as, whether you are moving the camera too fast or too slow, as well as visual feedback on what has and what hasn’t been captured, that’s going to drive adoption and as a result, the rate of advancement will only accelerate,” he says.

“With some of our younger team members, we’re able to give them a smartphone, give them 10 minutes of training and then they’re off doing it. Very quickly, people are going to recognise the value of this technology in a real-world setting. There’s going to be continuous improvements in terms of usability and scalability. It’s a no-brainer.”

“The costliest element of real-world data capture for projects is most often the human component, either in time to collect information or time to process it. The ideal future would be the creation of a digital model that’s indistinguishable from the real thing, that gets you to a point where you’ve collected enough information to never need to go back again.”

“To be able to collect all the data you need at once, then to have that perfect replica of the original, would be enormously valuable to us. It’d be gold.” 

This article was originally published in the April edition of our magazine. To read the magazine, click here.

Send this to a friend