Automation on the Docks Means Fewer Jobs — and Often No Improvement in Productivity
A new report finds that automation has led to significant job losses on West Coast docks. We spoke to an automation specialist and trade unionist who works at a fully automated terminal to see what that transformation looks like for those at the heart of it.
The Port of Long Beach in California. (Getty Images)
- Interview by
- Alex N. Press
Someone Else’s Ocean: Shipping and Jobs in the San Pedro Bay, a new report published by the Economic Roundtable and underwritten by the Coast Longshore Division of the International Longshore and Warehouse Union (ILWU), looks at a host of issues pertaining to the political economy of the West Coast’s ports, from cargo volume to the balance of trade and hours of dock work.
Considerable pages are devoted to automation, a subject at the center of ongoing contract negotiations between the ILWU, which represents twenty-two thousand workers at twenty-nine ports, and the Pacific Maritime Association (PMA), representing the shipping terminals.
Port automation entails the replacement of workers by machines that operate largely, though not entirely, autonomously. The issue has been at the heart of contract negotiations for decades: in 2008, the ILWU agreed to allow increased use of technology, and in the years since, the union has argued that automation has meant job losses, with automated terminals both requiring fewer workers and leading to less work at traditional terminals, as cargo shifts to automated terminals.
The PMA, for its part, insists on the opposite story.
“Automation allows greater densification at existing port terminals, enabling greater cargo throughput and continued cargo growth over time,” said Jim McKenna, the PMA’s chief executive, in a recent statement. The PMA recently put out a report arguing that not only does automation increase capacity, it also benefits labor. While the latter claim is widely disputed, so is the former, with several studies finding that automation does not increase productivity.
Four percent of global shipping capacity is currently automated, and that includes Long Beach Container Terminal (LBCT) at the Port of Long Beach and Trans Pacific Container Service Corporation (TraPac) at the Port of Los Angeles. Daniel Flaming and Patrick Burns, the authors of the Economic Roundtable report, find that automation reduced dock-work employment at LBCT by 37 to 42 percent, and at TraPac by 24 to 30 percent. At LCBT and TraPac, dockside employment was reduced by 37 to 52 percent and 34 to 37 percent, respectively. The authors write that a conservative estimate finds that automation eliminated 572 jobs annually at the two ports in 2020 and 2021.
To get a picture of what automation looks like for those subjected to the transformation, Jacobin’s Alex N. Press spoke with ILWU Local 63 member Rebecca Schlarb, an automation coordinator at LBCT, where she has worked for six years. Before that, she was a yard planner, and has worked in a variety of roles — super cargo, yard planning, rail planning, vessel planning — since she began dock work in 1991.
To start, tell me about your job.
I’m part of the Marine Clerks Local 63 in Los Angeles–Long Beach. I’ve been there since 1991, so I’ve seen a lot of growth and changes. Marine clerks are one-third of the portion of the labor that moves the cargo: we load the cargo on and off the ship, in and out of the yard, on and off the train. We plan out the yards and set them all up to help facilitate a smooth flow and transition of cargo through the port.
Given that you’ve worked at other terminals before, and now the automated terminal at Long Beach, how do those compare? What has changed in your tenure on the docks?
From the marine clerk perspective, technology has been affecting us since the 1980s, with the advent of computers. Whereas clerks receiving and delivering cargo used to do handwritten interchanges on each of the lanes — if there were twenty lanes, there would be twenty clerks out there — the advent of technology meant you could input through computer systems and databases, so instead of having twenty clerks down in the lane, now, in a remote tower, there’s six processing that same amount.
As far as the big top handlers that pick up containers and set them down within the yard, we’ve gotten GPS, so where one clerk used to be with each top handler, now maybe there’s one clerk for two, three, or four top handlers, because supposedly GPS is auto-spotting those containers in that position in the yard. That doesn’t always happen, so then the marine clerk comes through and updates the piles to make sure that the data is accurate within the system.
So I’ve seen that loss of jobs. Then, coming to LBCT, we have thirty-six full blocks with multiple bays. Two automatic-stacking cranes run up and down, so with those thirty-six blocks, there’s sxity-nine cranes within each of those blocks for yard delivery. In a traditional operation, that would have required 138 crane operators and then sixty-nine signal people, who are the safety people down on the ground communicating with the crane operator. Instead of 138 crane operators, now, in a remote office, there’s only fourteen crane operators and those sixty-nine signal people are eliminated. That’s a substantial loss.
As for marine clerks, on a traditional operation, we would have had twenty-five to thirty clerks working with just yard delivery, and that’s been knocked down to four. Against the ship, we used to have the semitruck drivers that would pull the bomb carts that containers are set on and moved about to and from the crane to the yard. What we have now are automated ground vehicles (AGVs) that move that cargo in a safe restricted area. We had about eighty-five of them working, which means eighty-five drivers who have been cut, so now we have maybe two drivers against the vessel. As far as floor runners that would have been against the ship, we would have had fifteen or more, depending on how much equipment was moving and what needed to be spotted, and that’s been reduced to two.
As the terminal has grown, we did pick up seven very specialized automation coordinator jobs, which is the position I now hold. I sit at a desk with six monitors, looking at multiple programs and cameras to facilitate the five transition points to and from the ship. So that would be a work instruction for the ship-to-shore crane going to and from the crane to a platform area where the swing men take on and take off the cones, and there’s a key dock person, a safety person, for that area.
The crane operator sets these cans down from the ship on the platform. Then, from the platform, a portal trolley comes in and picks up the container off the platform and drops it down on this AGV. Then that AGV drives to the block and those auto-stacking cranes out in the yard pick it up. So I monitor those five transition points to make sure that no exceptions are happening. I work with the crane operator, dock foreman, crane mechanics, special mechanics for the AGVs, and the reefer mechanics.
This new study estimates that 572 full-time jobs have been lost at LCBT and TraPac annually in the past two years.
I’d say that’s at minimum. Maersk is in the process of transitioning into something of an automated terminal as well, and I anticipate that will probably cut out about 300 to 350 jobs a shift, which means 900 jobs a day potentially lost. There are rumblings of another terminal automating similar to LBCT, so that would be substantial too.
What do you see as a solution to the push toward automation?
My biggest frustration is that automation is not that efficient. I was finishing a vessel last night and, of the thirty-six blocks, at one time a bunch of blocks went down, because every time a crane breaks down, both cranes have to be taken out of operation along with the two cranes adjacent to it so that it’s safe for the mechanics to drive down to do the repairs. That stops hundreds, if not thousands, of cargo containers from being moved for anywhere from twenty minutes to an hour. The truck drivers can’t get their cargo, I can’t get it for my vessel, the rail can’t get it for their train from that block.
As time goes on, I’m seeing all types of automated equipment breaking down, the ship-to-shore cranes over the vessel, the automated stacking cranes (ASCs), as well as AGVs. This downtime can be as short as six minutes or as long as an eight-hour shift or a few days to repair. For the ASC in the yard, the average minimum downtime is twenty minutes to an hour, and that has the greatest impact for the entire terminal. If one ASC goes down in a block, it is not only the broken crane that’s taken out of operation but the second crane within that block as well.
Additionally, for the safety of the mechanics, both ASCs in the adjacent yard block (a total of four ASCs) are also taken out of operation to provide safe passage to the broken crane. This means that none of the hundreds of containers stacked in either block can be moved while repairs are made. No deliveries to outside truck drivers, no cargo to the rail, or to the vessel from the affected yard blocks. If it’s determined the broken ASC can’t be repaired, it can be parked out of the way, allowing the remaining one ASC to service the entire yard block for all three operations. This substantially limits the volume of cargo that can be moved for the remainder of the shift from that yard block. It is the outside truck driver that suffers the most from these mechanical breakdowns, though it can also be a challenge for the vessel or rail to continue to load cargo if a suitable replacement from a working yard block is not available. It has been my experience during a shift that up to six blocks may be down off and on throughout the shift, with an additional five or six blocks with only one ASC for the entire shift.
In regard to AGVs, they are the main source of cargo movement for the vessel, and are also subject to mechanical issues that can affect vessel production. There are multiple reasons why an AGV may break down, but the most difficult error to correct is due to a loss of heartbeat, our term for the wireless network connection. An AGV follows a travel path, generated by the program software, from one part of the yard to another. When the wireless connection is lost, the AGV immediately stops driving, and a large safety block is generated by the software, stopping all other AGVs in close proximity.
Because the AGV with the lost wireless connection could be physically positioned anywhere within the anticipated travel path, it may not be possible to reroute the adjoining AGVs so they can continue working. If this occurs, a large segment of the field may need to be stopped by a safety net to allow the mechanics to physically enter and manually reset and reposition the broken AGV so the wireless connection can be reestablished. This process may take fifteen to thirty minutes to perform.
I mention all this intricate information, which is actually a small sample of all the problems I face each shift, to provide background for understanding why a conventional labor-driven operation is more productive than automation. Let’s start with the AGVs: if a car or truck breaks down in front of you, all that is required is that you reverse course and visualize a safe path to proceed around the obstruction, taking just a few seconds to a few minutes to execute. We as humans are not limited to a linear line of programming code. An experienced longshoreman can see a solution to the problem in a fluid and ever-changing landscape. In regard to the yard block, where hundreds of containers are decked within the row, if a piece of equipment breaks down in a conventional labor operation, only the one, or possibly two, individual bays of twenty to thirty containers are blocked, not two entire rows of hundreds of containers.
A conventional ILWU longshore labor operation has repeatedly proven to be more flexible and productive than automation. When the unexpected happens, we as humans are able to adapt and utilize additional physical places within the terminal to manage cargo. Automation cannot do this without months of recoding parameters to accommodate this change and updating GPS positioning. The current massive influx of cargo volume we are experiencing in the ports of LA–Long Beach is a perfect example of the human mind’s ability to adapt and utilize every potential space in the port. It is important to note that over the last decade and a half, cargo volume through both ports has attained record-breaking numbers. This is not due to automation; it is due to the members of the ILWU who have worked incredibly hard to keep our part of the supply chain moving.
There’s nothing more beautiful than watching a gang of longshoremen executing a plan, watching it flow and move without a hitch, making adjustments, and so on. Nothing moves better than actual hands-on interaction. The only positive argument for automation is it saves labor costs for shipping companies and terminal operators to rake in billions of dollars in profit that does not stay here in the United States. They’re the ones driving the terminals to go automated, so they can earn even more money and really devastate the community in the process, and the money our wages translates to in terms of a tax base and stores in our areas.
Is there anything else people should know about automation on the docks?
Ours is a fascinating job, and it’s a dangerous job. We’ve had other union brothers and sisters killed on the job as well as seriously maimed and unable to continue on, to make sure that everybody gets their goods. It’s dangerous work, but it’s a beautiful thing to watch and we all work really hard.
There’s a perception that we’re overpaid, we’re lazy. But to get people their goods, we do a great job, even with automation, and what we do supports everything down the line across the country.
I also think every American should have what we have: an excellent health care plan and a pension that, after they’ve worked thirty to forty years in an industry, allows them to maintain a dignified life afterward. It’s sad that so many Americans don’t have that.