The Amazing Ways Artificial Intelligence (AI) Can Now Detect Dangers At Work
More people die in construction than in any other industry, and the number one cause of death on a job site is falling.
Autodesk’s latest addition to its BIM 360 suite of artificial intelligence (AI) enabled industry tools – Construction IQ – aims to reduce these tragic occurrences. It does this by predicting when falls are likely to happen – as well as any other danger to life, limb, or even just quality of work.
Autodesk’s data scientists hit upon the solution while looking for applications where the massive amount of data collected on modern-day construction sites could be put to use, thanks to the industry’s enthusiastic adoption of mobile tools and sensing devices.
“Imagine being a construction manager and having to contend with the fact that every X number of months, someone’s going to die on the job – it’s unfathomable to most of us in white collar jobs,” says Pat Keaney, Autodesk’s lead on the Construction IQ project.
Construction IQ uses natural language processing (NLP) techniques – algorithms that parse human language (in this case, text notes created around construction jobs by contractors and subcontractors on site) to assess risk and warn of hazards that may go unnoticed by human safety managers.
Keaney tells me, “Right now we’re focussing on applying NLP … we have partners using image recognition; there are 360-degree cameras, Internet of Things (IoT) that can detect gasses in the air … in my mind there’s no doubt that within the next five years, this technology is going to be saving lives.”
In fact, evidence suggests that it probably already is. I spoke to one construction company, BAM Ireland, which told me that by using Autodesk’s BIM 360 platform, they had achieved a 20% reduction in quality and safety issues on site.
They have also increased the amount of time available to staff to spend remedying high-risk dangers on site by 25%.
All of this has become possible thanks to the explosion in the amount of data generated and gathered at construction sites.
“We realized there are these tremendous changes happening in the industry as a result of things like smartphones and tablets,” Keaney says.
“They’ve only been around 11 or so years, but they’ve changed the landscape … instead of carrying around reams of paper, you can look at plans on an iPad, and when you do this, you don’t just save time, you generate and collect data.
“Everyone has a phone and a high-definition camera in their pocket … so expand that idea to IoT and sensors, most smart people don’t sit around trying to figure out when concrete has cured now, they put a sensor in it, and the sensor tells them when it’s ready.”
The next stage was a natural leap – taking this data and making it available to the AI tools Autodesk developed for its BIM 360 platform meant a move towards a predictive, data-driven model of construction management.
“Our simple hypothesis was there’s got to be value in this data that will help our customers do a more effective job of managing their crazy, chaotic, every-growing construction projects. That’s what led us to start this exploration.”
This is a perfect example of an increasingly common and productive trend across all industries that are engaging in digital transformation. Digitization results in a wealth of information that can often prove useful far beyond the initial use cases for which it was collected.
As the project got underway, there were initial concerns about how willing companies would be to share the data. These proved to be unfounded, as Keaney explained to me:
“In general, our customers have far exceeded my expectations for willingness and passion for allowing us to help them find value in their data … we said we want to go on an exploration and partner with you guys … if you’re interested in what we need is for you to grant us access to your data.
“When we did that, our customers would get really excited and dig in and want to spend more time with us … we were able to show them things in their data they had never seen before.
“Are there certain companies that were more conservative and wary? Certainly – in the US they were more willing to take a risk, Europe was a little more cautious – which is one of the reasons it was so exciting to see it embraced by BAM Ireland.”
Data covering over 150 million construction issues harvested from 30,000 real-world projects has been used to train the algorithms that BAM used to drive their impressive results in the field of site safety.
Their digital construction operation manager, Michael Murphy, told me how the platform had allowed them to move away from the siloed approach to data the construction and civil engineering firm had traditionally taken.
He said “When we started, we found our biggest problem was our data was very inconsistent, so when we were setting up projects we were being inconsistent around how we were capturing data, or the issue types we were capturing.
“When we engaged with Construction IQ, the first thing we had to do was tackle this inconsistency – that was a big lesson learned.
“This meant we were able to get better insights into the issues and challenges across our projects … whereas previously we may have just been doing something on a mobile phone or an iPad for the sake of doing it on an iPad … we weren’t really getting the benefits of having standardized datasets that we could query, and get better insights from.”
It seems inevitable that as technologies such as machine learning, NLPand deep learning continue to prove their worth, solutions built on them will become increasingly widely adopted across construction, as well as any other industries that can benefit from a consolidated approach to data gathering and analytics.
In the short term, this is likely to save lives, while in the long term, it will contribute to the development of safer working practices and standards.
As Keaney puts it, “I think safety is something that everyone can agree is important – nobody should be holding their data close to their chest around these issues, it wouldn’t be good behavior.
“The whole industry shares these problems … and all of this tech is going to save lives; there’s no doubt about it … from a safety perspective, the benefits are clear, compelling, and really easy to understand.”