AI Resilience Report forStructural Iron and Steel Workers

In steel construction today, most core tasks are still done by people. For example, workers read blueprints and guide cranes by hand. Computers and tools are helping more with planning: companies use 3D design software (BIM) to model structures before building ^[1].

Some new technologies are being tried on jobsites. One university project used augmented-reality glasses to overlay shop drawings on steel, showing fabricators exactly where to cut or bolt pieces ^[2]. 3D laser scanners can also capture a steel frame as it’s built, and software compares it against the digital design to find misalignments ^[3]. In research labs, engineers have built mobile welding robots that use cameras and AI to find and weld steel joints automatically ^[4].

Others are developing wearable exoskeleton suits that give workers extra lift, letting them carry heavy beams more easily ^[4]. However, these tools are mostly in testing or factory settings. Daily on-site work – hoisting beams, bolting connections, checking levels – is still done by ironworkers.

As one industry review notes, every construction site is unique and constantly changing, so on-site automation is still rare ^[1]. In summary, technology like computer models, lasers, and even smart tools is starting to help, but the tricky physical work of fitting and finishing steel is still done largely by people.

Ironworkers face strong demand and hard labor. About 85,000 ironworkers (structural and reinforcing) worked in the U.S., with job growth around 4% ^[5]. The work is dangerous and demanding (working at heights, heavy lifting), so there is interest in tech that can help.

For instance, robots can do repetitive or heavy tasks faster and with less injury risk ^[6]. A shortage of welders (over 330,000 more welders expected to be needed by 2028 ^[4]) motivates research on welding automation. But bringing AI onto the jobsite is costly and complex.

Robots and sensors are expensive investments, and a skilled crew can be cheaper for many jobs. For example, the average ironworker earns about \$32.80 per hour ^[5], so a new machine must really save on time or mistakes to pay off. Construction sites also change every day – weather, layouts, and people keep moving – which makes running robots safely very hard ^{[1] [6]}.

Because of rules and safety concerns, companies adopt new tech slowly. Some contractors use drones or scanners to inspect and plan, but most fitting and bolting is done by seasoned workers. Experts advise that future robots will more likely assist ironworkers rather than replace them ^[1].