Artificial Intelligence in Construction Equipment
By Gari Nickson, Construction Executive
How is artificial intelligence (AI) being used to make construction equipment safer and more efficient? Construction has always been a dangerous business and not always the most efficient. With heavy machinery involved, often uneven terrain, constant activity and the risk of human error, the threat of danger is never far away. And despite the ever-increasing focus on health and safety and use of technology on jobsites, there are still plenty of fatalities.
In the U.S., the number of deaths on site has risen 34 percent since 2010. In Japan around 300 deaths and more than 15,000 injuries were recorded in 2016, as reported by the Japan Construction Occupational Safety and Health Association. In the U.K., however, things are improving. By the end of March 2017, fatalities had reached all all-time low. Nevertheless, there were still 30 deaths on building sites in the 12 months, which is still well over the all-industry average of 0.43. But one death is one too many.
However, thanks to AI, companies are realizing the benefits that machines can bring, tackling not only health and safety issues in construction, but making work more efficient. The industry appears particularly well-suited to being able to be improved by the wonders of AI.
One interesting example is a tie-up between Japan’s Komatsu, one of the world’s largest manufacturers of construction and mining equipment, and U.S. tech company NVIDIA. The two firms are working together to use NVIDIA’s graphics processing units – essentially intelligent cameras – to visualize and analyze entire construction sites.
Komatsu will create 3D images of sites, revealing the real time movement of machinery, people and objects. Equipment onsite can also be monitored to make sure it is utilized with optimum efficiency.
NVIDIA’s smart cameras will also communicate with cameras and drones on the jobsite, to act as an analysis and visualization AI platform.
Technology called SkyCatch will allow drones to map and gather 3D representations for visualizing the terrain at the edge. OPTiM (Iot management-software firm) will also offer an application to recognize machinery and workers collected from surveillance cameras.
The centerpiece of the tie-up is NVIDIA’s Jetson. Working alongside cloud technology, this platform will power cameras on construction equipment to allow 360-degree views, meaning machines and people nearby can be seen to avoid collisions and possible accidents. The cameras will also recognize fast-changing conditions and inform workers in cabs to act accordingly.
In the future, other applications planned include virtual simulations and high-resolution rendering of construction sites as well as automated machinery control.
Elsewhere, construction firms are also taking advantage of new AI technology to help avoid potential accidents. Arup and Skanska are among the companies using a video and photo platform called smartvid.io. This technology implements AI to sort through large amounts of videos and images captured on construction sites to identity possible hazards.
Now this is the clever bit: the software utilizes algorithms of image recognition to spot particular search criteria such as safety vests, hi-vis colors and hard hats so it can uncover images that will highlight workers who aren’t kitted out in the correct safety gear or acting against regulations.
In the space of minutes, results are collated, collected and sent to a site manager. This may feel a bit Big Brother-like for workers on the jobsite. But for a manager it’s a dream, as it’s a job that could normally take hours to do.
In a similar vein, Swedish firm Volvo Construction Equipment is also looking at new technologies to boost the workers safety on building sites. The company is building AI algorithms that decipher and detect particular objects using computer vision techniques. The system then transmits a warning to the operator in order to cut the hazard of a potential accident.
AI has plenty of other benefits, such as performing repetitive, mundane and often dangerous jobs. “AI techniques can help inexperienced machine operators to carry out complex tasks, which they otherwise could not. It’s advantageous in an industry that’s finding it increasingly difficult to source highly skilled and experienced operators,” says Dr. Fares Beainy, machine intelligence program leader at Volvo Construction Equipment.
Indeed, speculation is mounting that autonomous equipment in heavy industry could become much more common well before self-driving cars are widely adopted.
Currently full automation is generally only used for heavy good vehicles that must make repetitive movements, such as hauling ore. So what are the reasons that autonomous vehicles could become more common? A key reason relates to productivity, very much a business buzz word at the moment, but particularly in the construction industry. Operating heavy equipment means employing somebody with specialist skills, whether it be controlling a crane or an industrial earth-moving vehicle. These types of skills are not always easy to find and as such can be costly. Therefore firms are likely to be open to the idea of adopting autonomous vehicles for this type of work. A second reason is that clearly self-driving cars will have plenty of obstacles to navigate, be it a distracted pedestrian, a stray cat or confusing traffic lights. However, on a building site there are clearly are less unpredictable obstacles, so an autonomous vehicle would have less to cope with.
All these developments reveal that construction is ripe for embracing the latest opportunities that AI offers when it comes to improving how equipment is used. It helps firms and industry tackles fears around health and safety and boost productivity – a real win-win scenario.