Using sonar and A.I. to detect cracks and structural damages. Startup in Residence Northern Netherlands participant oQuay is helping the province of Friesland to better inspect the many banks and quays along its waterways. We catch up with founder Peter Nieuwenhuizen to discuss the progress so far, the challenges and the near future.
oQuay was founded in 2019, as a way to tackle one of the biggest issues of waterway inspections: getting the full picture. “I’ve been doing this type of work for 12 years and, whether you’re inspecting the structural integrity of bridges, flood gates, tidal locks or quays, the biggest issue is trying to figure out what’s below the waterline”, Peter explains. “Even if you use divers equipped with lights and cameras, there’s still limited visibility and you also need to combine and synchronize that information with what’s going on above the waterline. So that’s why we started oQuay, using a combination of sonar and videomapping to get the full picture.”
The Province of Fryslân owns around 220 kilometers of banks and quays along important waterways. And all of those banks and quays need to be inspected periodically, of course. As you can imagine, these inspections are a very labour intensive process. And it’s also done by people, meaning there is a risk of different interpretations of the overall condition of the bank in terms of maintenance and safety, which reduces the quality of the inspections.
Sounds like the perfect challenge for oQuay. “We’re talking about 200 kilometers, which is a ton of data to process manually. So we asked our partner Merlakor to use A.I. and train the software to automatically recognize and pinpoint cracks and structural damages”, Peter explains. “This would save a lot of time and effort and we’re currently running a pilot to test and train the algorithm. We’ve just finished the first round of testing and are getting ready for the second round.”
Speaking the same language
“It’s going pretty great so far”, Peter says. “We have weekly meetings with the Province to discuss the progress and what still needs to be done. The tricky thing about our line of work is that it’s not an exact science. It’s not about being able to spot damage, it’s really about assessing and defining the extent of the damage and also agreeing on that before you can take the next step.”
“It’s like having a dent in your car”, Peter continues. “You could say, well, that’s damage, so it’s a problem. Or you can say the car still works, so it’s not ‘real’ damage and not a problem. Both statements are equally true, but have different definitions and opposite outcomes. The Province speaks our language in that sense, so that really helps. And the Province of Groningen is now also on board, which is great news too.”
So far so good, but they’re not there yet, Peter says: “Right now, the software can determine if there’s damage, but without a value judgement. But the next question is: how severe is the damage? Does it need fixing? How much is that going to cost? And is it worth the cost? So for us, this is just step one of a much bigger process.”
A.I. aside, oQuay has another exciting ambition for the next few years: “Underwater drones are becoming more popular, so we’re also busy expanding on that front. Hopefully in the next few years, we’ll be one of the top companies in the Netherlands in that area, because our drones are up to that specific task, we have the right sensors and experts to pilot them, but who also have the experience with these types of inspections and know what to look for.”