Automated cargo-haulers towering four-stories high glide among the waiting boxes, straddling and lifting them before wheeling them to the nearby “stacks.” Here the boxes are passed off to another massive robot—an automated stacking crane—that arranges them into meticulous stacks. When it comes time for a specific box to continue its journey inland, those same robotic cranes will find it and load it onto a waiting truck—no human operator necessary.

TraPac terminal—along with a terminal at the nearby Port of Long Beach—is among the first U.S. ports experimenting with robots, artificial intelligence, and other digital tools to choreograph the complicated dance that keeps goods flowing into and out of major U.S. ports. The technology—though not without its critics—is widely seen as the most efficient way for seaports to cope with rising global shipping traffic and massive new ships that haul more and more containers. By digitizing and automating activities once handled by human crane operators and cargo haulers, seaports can reduce the amount of time ships sit in port and otherwise boost port productivity by up to 30% by some estimates.

The automated facilities at the ports of Los Angeles and Long Beach—two of the nation’s busiest—are important proving grounds for technologies that have firmly taken root in European and Asian seaports but remains a relative rarity in the U.S. Only four U.S. seaport terminals currently use the technology. The other two, in Virginia and New Jersey, were the first in the U.S. to implement dockside automation. But while some of the world’s largest container ships make calls at East Coast docks, they rarely unload all of their cargo at a single port as they do on the West Coast.

That means West Coast shipping terminals are likely to automate faster than their East Coast counterparts, placing the ports of Los Angeles and Long Beach at the front of a wave of automation needed to bring U.S. shipping logistics up to speed.

Doing so is not as easy as taking one facility’s automated systems and grafting them onto the next terminal, Port of Los Angeles executive director Gene Seroka says. “You’ve got probably 190 or 200 ports in the United States, you’ve got 25 ports of national significance as classified by the Army Corp of Engineers. Each one is different. Once you’ve seen one port, you’ve still only seen one port.”

Still, there are common components at every seaport where automation can make the work more efficient. Every seaport needs cranes to load and unload cargo, for instance, as well as a means of moving containers around the storage yard in an organized fashion so that specific boxes can be located at the specific time they are needed. These kinds of coordinating, organizing, and choreographing tasks are more efficiently handled by machines than by humans. It doesn’t hurt that robots and algorithms don’t require breaks, weekends, or health insurance.

For terminal operators, automation isn’t just about handling more cargo. Automated systems also allow seaports to boost the efficiency of one of their most limiting, finite resources: space. With only so much waterfront property to go around and the volume of cargo rising, seaports face a dwindling amount of real estate in which containers can be stacked and stored, even if only for a few hours.

“The real opportunity is densification,” says Dr. Asaf Ashar, an emeritus research professor at the University of New Orleans’ National Ports & Waterways Institute and an independent consultant for the shipping and transportation industries. “You can use existing land more efficiently.”

But automating seaports is also wildly expensive. The process doesn’t consist of any single thing, but a continuum of digital technologies, software systems, and robotic hardware. Deploying automation to any given port terminal can cost more than $2 million per acre, Seroka says. Upgrades to the Port of Long Beach’s automated terminal will cost in excess of $1.3 billion by the time the technology is all in place in over two years.

Automating the 210-acre TraPac terminal at the Port of Los Angeles will likewise cost more than $1 billion in public and private funds, and it’s not exactly clear when or if that investment will pay off. APM Terminals, a part of Dutch shipping giant A.P. Moeller-Maersk, told the Wall Street Journal that its automated terminal in Rotterdam—where terminals have embraced various levels of automation going back to the 1990s—has paid dividends, requiring just half the human labor that its conventional terminal at the same port requires. But when the same company opened the first semi-automated port in North America in Virginia in 2007, similar returns didn’t immediately materialize (APM Terminals has since sold the facility).

The high costs and shaky record of returns—not to mention pushback from labor unions intent on preserving seaport jobs—have left stakeholders in other U.S. ports leery of going all-in on dockside automation, allowing ports in Europe and Asia to become models of modern logistics. That in turn makes the experiments at Long Beach and Los Angeles all the more critical. The Ports of Los Angeles and Long Beach are the No. 1 and No. 2 for container volume in the United States, respectively. In 2015, a total of 8.16 million TEU—the equivalent of 20-foot containers—moved into Los Angeles, while another 7.19 million TEU came through Long Beach, according to the World Shipping Council.

“Automation is a business decision and it really has to pencil out before any further work is pursued,” LA’s Seroka says. Making the math “pencil out” should grow easier as other digital technologies—artificial intelligence and big data analytics among them—are further integrated into the seaport enterprise as well. A pilot program currently underway at the Port of Los Angeles with GE—known as the Port Information Portal—digitizes maritime data and will ultimately create a computer dashboard that provides a window into the entire port supply chain. The test was recently extended to run for three years—a testament to the support the technology has garnered at the port, including from organized labor.

“With respect to digital technology, for cents on the dollar we can expand this port capacity,” Seroka says. “We can do it quicker than waiting for the current administration to work out an infrastructure plan, utilizing the land that is within inside our four fences, almost immediately.”

Clay Dillow is a freelance journalist. Brooks Rainwater is senior executive and director of the National League of Cities’ Center for City Solutions.