7 ms·

A Tale of Two Ports: Automation at Oakland vs. Rotterdam

68 points by Bootvis 11y ago

JabavuAdams 11y ago
TIL the average salary for port workers is $147k.
- fennecfoxen 11y ago
  With benefits worth another $82,000 per year! They don't need to deduct from that salary for a 401(k), that's certain.
- Frondo 11y ago
  That's the kind of quality of life a strong union will get you as a worker. No wonder the biggest business interests have undergone a decades-long propaganda campaign against them.
  - enraged_camel 11y ago
    >>That's the kind of quality of life a strong union will get you as a worker. The flip side of the coin is that the quality of air you will breathe every single day is guaranteed to be abysmal, and air quality has a very strong inverse correlation with life span.
    - JabavuAdams 11y ago
      Why's the air so bad? Because of the bunker oil used on ships, or some other reason?
- tedks 11y ago
  This is the link in that sentence: http://www.pmanet.org/the-ilwu-workforce http://www.pmanet.org/the-ilwu-workforce Clicking around on that site shows that it is a non-profit representing the companies that negotiate with the longshoreman's union: http://www.pmanet.org/ceo-message http://www.pmanet.org/ceo-message So it's the bosses' word that the average salary for port workers is 147k, and honestly I don't believe it.
- bsder 11y ago
  And what's the separation between the mean and the median? One worker making $1,000,000 and 9 making $10,000 gives you an average of $109,000 while only a single worker is above the mean.
troisx 11y ago
I can't wait for more automation to hit US ports. I'm all for decent wages, but dockworkers get paid so much they end up behaving a lot like oilfield workers. Dockworkers are a prime example of what gives unions a bad name.
- kalkin 11y ago
  Do you think nobody should be making six-figure salaries? Or just that unionized workers shouldn't?
  - compostor42 11y ago
    The idea of blue-collar workers getting well-paid is very upsetting to some white-collar workers. They feel their social status threatened by people they deem beneath them.
    - lambdapie 11y ago
      I don't think that's the primary issue people have with high wage union labor, and your comment shows an unwillingness to engage constructively with political opponents. Unions are a monopoly, and when prices (in this case, of labor) are high due to a monopoly, we must ask whether the government should allow, forbid, or encourage this monopoly. When wages are very low, I think most people are not opposed to unions forcing wages above market rates (I am, but that's another issue). But when wages are high, or driven far above what people think market wages would be, people question the moral basis for protecting workers from competition. So the issue isn't that blue collar workers are "beneath" white collar workers, but that (at least to an industry outsider) it's not clear what special skills these workers have that would command high market wages.
      - ConroyBumpus 11y ago
        My issue with union shops is that getting things done is unnecessarily difficult. Case in point: sitting around for half a day waiting for the unionized cable puller to sling a single cat5 connection between two racks on the same row in a data center. Then the guy doesn't show up until after 1600, so he forces overtime.
SixSigma 11y ago
The real automation in the Rotterdam port is the Automatic Guided Vehicle system that moves containers between the loading cranes and a container re-arranging system that optimizes the placement of containers dock side. The loading cranes themselves are only remote control on the deep water side, and this is for health and safety reasons. The cranes for the trains and barges for the hinterland are sit-in cabs. SOURCE: I did a tour of the terminal a few weeks ago as part of the 6th International Conference on Computational Logistics [1] A similar AVG system is in place at Hamburg as seen in this video [2] [1] http://realtimelogistics.info/iccl/ http://realtimelogistics.info/iccl/ [2] https://www.youtube.com/watch?v=3B3PzgYUK_0 https://www.youtube.com/watch?v=3B3PzgYUK_0
- DenisM 11y ago
  This video suggests the cranes are automated as well: https://www.youtube.com/watch?v=WxXZQ7emHC0 https://www.youtube.com/watch?v=WxXZQ7emHC0
Animats 11y ago
TraPac and Middle Harbor in the Port of Los Angeles are automating. Oakland isn't considered big enough to need it.[1] for an overview of automation in Los Angeles ports, see [2]. The Port of Los Angeles isn't a port operator; it's a landlord. Thirteen private companies have their own terminals, docks and equipment, and two of them are automating now. To automate a terminal, you have to take it out of service and build a new one. The above references may be too specialized for general readers. Here's a "gee whiz" video of Rotterdam port operations.[3] And a tutorial with animations, from the port of Busan, S. Korea.[4] Note that the article is partly an ad for FlexPort. Automated ports have less need for FlexPort's service. Most automated ports have buffers, big stacks of containers. Containers come off ships and are transferred to container stacks nearby. On the other side of the stacks, trucks approach and have containers loaded onto them. The terminal handles container sorting. Oakland loads many containers directly from ships to trucks towing container trailers, without buffering. Matching up trucks and containers is a problem. FlexPort's concept is "Uber for containers" (I know, but here it's real.) Any truck, any container, any destination within some distance limit. If Oakland had automated container stack buffers, FlexPort wouldn't be needed. [1] [http://www.joc.com/port-news/port-productivity/us-ports-weigh-value-terminal-automation-investment_20141002.html http://www.joc.com/port-news/port-productivity/us-ports-weig...] [2] https://www.portoflosangeles.org/Board/2014/April/042014_item5_Transmittal_3.pdf https://www.portoflosangeles.org/Board/2014/April/042014_ite... [3] https://www.youtube.com/watch?v=WxXZQ7emHC0 https://www.youtube.com/watch?v=WxXZQ7emHC0 [4] https://www.youtube.com/watch?v=2t2cqiQl0BE https://www.youtube.com/watch?v=2t2cqiQl0BE
- fraserharris 11y ago
  FlexPort's customer is the business paying for the shipping (freight forwarder & customs broker), not the port. Their interest in port automation* is that it makes for interesting content marketing and it may benefit their end-customer someday. * from what I can tell
Maarten88 11y ago
Rotterdam also has unions to deal with. A colleague of mine, who worked on the software for the final stage (some 15 years back) moving containers onto trucks, told a story where, after union negotiations, they implemented a system where forklift truck drivers could choose between three different containers to pick next instead of giving them just one instruction, for the single, most optimal, one.
- macmac 11y ago
  While the unions are arguably stronger on the US East Cost the International Longshore and Warehouse Union at the Oakland terminal certainly isn't a pushover.
- cosmolev 11y ago
  What is the logic behind it? Having an opportunity to select is more entertaining to the driver?
  - chime 11y ago
    I am in the middle of writing directed pick/put module for a warehouse. I just had a meeting last week on the exact same topic. My software gives the user the most optimal pick list and the warehouse manager wanted me to change it so his team could pick from alternate bins. His justifications: 1. My software doesn't know that there is a cart/ladder/forklift blocking the most optimal bin while the bin next to it is easily accessible. 2. My software cannot tell if QC is in middle of sampling or if Inventory team is in middle of cycle counting that exact bin. 3. My software doesn't factor in changes in real-world that have not yet been inputted into the computer system. So if there is spillage, quantity, or quarantine issues that haven't been scanned into the system yet, my software calculates the optimal bin incorrectly. It's not that code cannot calculate what optimal means given a set of requirements. It's just that the end-user might be in a better position to decide if it really is optimal or not.
kalkin 11y ago
This article tells you basically nothing if you're not already convinced that more automation is always better, or that while "technological disruption... almost inevitably leads to displacement of some segments of the workforce. In the long run, however, technology ends up creating better jobs and expanded opportunities across broad spectrums of the economy." The calculation that "proves" Rotterdam is more efficient is based on the unargued and generally false assumption that both ports are operating at full capacity - the author just divides TEUs by number of cranes. TEUs handled per year in Oakland only just matched their 2006 peak after falling drastically during the Great Recession: http://www.portofoakland.com/maritime/containerstats.aspx http://www.portofoakland.com/maritime/containerstats.aspx And Oakland's main capacity limitation actually seems to be congestion getting containers out of the Port: http://www.wsj.com/articles/oakland-port-proposes-trucking-fees-to-fight-congestion-1440103747 http://www.wsj.com/articles/oakland-port-proposes-trucking-f...
- tajen 11y ago
  I was also expecting hard figures about how many containers and shifts have been lost in accidents compared to an automated port.
twic 11y ago
How is there an article about the costs of using containers without automated infrastructure on Hacker News and nobody's made a Docker/Kubernetes joke yet?
- lgas 11y ago
  Why didn't you do it?
vool 11y ago
''But in many ways, the Port of Oakland, and most American ports in general, are some of the most technologically antiquated in the world.'' This BLDGBLOG post from earlier in the year details to automation on the on the East coast. http://bldgblog.blogspot.ie/2015/04/infrastructure-as-processional-space.html http://bldgblog.blogspot.ie/2015/04/infrastructure-as-proces...
Someone 11y ago
Earlier discussion (different article, but same subject): https://news.ycombinator.com/item?id=10306204 https://news.ycombinator.com/item?id=10306204 (Other article at http://priceonomics.com/why-arent-americas-shipping-ports-automated/ http://priceonomics.com/why-arent-americas-shipping-ports-au...)
joecomotion 11y ago
IIRC, Rotterdam had a massive govt. subsidy for going fully-automated, which kind of helps. I spent 10 years between 1997 and 2009 (took a break for grad school) working on semi-automated tracking systems for ports up and down the West Coast (and other places), including one in Oakland. The recent surge in interest in intermodal container ports sure brings back memories! My company developed niche systems for tracking container location within the terminal by instrumenting and tracking any crane/truck/etc. that could move a container. Even though longshoremen were still operating the equipment, it helped to have a real-time (and historical) picture of what happened, because planning systems were now less subject to clerical mistakes. Here's their (sort of terrible) website: http://www.nowsol.com/ http://www.nowsol.com/ These guys pretty much own the market for the terminal-planning systems we fed: http://navis.com/ http://navis.com/
remarkEon 11y ago
From the Wikipedia page, it looks like Rotterdam had 36,315 vessel arrivals in 2008 while Oakland had only 1,928. Is this comparison really apples to apples? They point out the difference in tonnage, but I'm just wondering if there isn't a better comparison to be made. https://en.wikipedia.org/wiki/Port_of_Rotterdam https://en.wikipedia.org/wiki/Port_of_Rotterdam https://en.wikipedia.org/wiki/Port_of_Oakland https://en.wikipedia.org/wiki/Port_of_Oakland
- jpatokal 11y ago
  As noted in the article, Oakland is containers only while Rotterdam handles other types of cargo as well, and their comparison is based purely on containers. However, all things being equal, it should be more efficient to unload a few ships with lots of cargo compared to lots of ships with little cargo. That said, Oakland is somewhat weird choice, it's the world's 62nd busiest (by TEU) when Rotterdam is 11th. LA is the busiest in the US at 18th, with Long Beach (also greater LA) not far behind. https://en.wikipedia.org/wiki/List_of_world%27s_busiest_container_ports https://en.wikipedia.org/wiki/List_of_world%27s_busiest_cont...
exar0815 11y ago
This basically only shows again what the tiny Netherlands can achieve. I dont think there is a single country that threatened by a force of Nature. But instead of giving up, you basically become the best Worldwide in construction in, on and around the water. Basically, if you have any problems with water at any Kind of level, just ASK the dutch.

68 points by Bootvis 11y ago

JabavuAdams 11y ago

TIL the average salary for port workers is $147k.

fennecfoxen 11y ago
With benefits worth another $82,000 per year! They don't need to deduct from that salary for a 401(k), that's certain.
Frondo 11y ago
That's the kind of quality of life a strong union will get you as a worker. No wonder the biggest business interests have undergone a decades-long propaganda campaign against them.
- enraged_camel 11y ago
  >>That's the kind of quality of life a strong union will get you as a worker. The flip side of the coin is that the quality of air you will breathe every single day is guaranteed to be abysmal, and air quality has a very strong inverse correlation with life span.
  - JabavuAdams 11y ago
    Why's the air so bad? Because of the bunker oil used on ships, or some other reason?
tedks 11y ago
This is the link in that sentence: http://www.pmanet.org/the-ilwu-workforce http://www.pmanet.org/the-ilwu-workforce Clicking around on that site shows that it is a non-profit representing the companies that negotiate with the longshoreman's union: http://www.pmanet.org/ceo-message http://www.pmanet.org/ceo-message So it's the bosses' word that the average salary for port workers is 147k, and honestly I don't believe it.
bsder 11y ago
And what's the separation between the mean and the median? One worker making $1,000,000 and 9 making $10,000 gives you an average of $109,000 while only a single worker is above the mean.

troisx 11y ago

I can't wait for more automation to hit US ports. I'm all for decent wages, but dockworkers get paid so much they end up behaving a lot like oilfield workers. Dockworkers are a prime example of what gives unions a bad name.

kalkin 11y ago
Do you think nobody should be making six-figure salaries? Or just that unionized workers shouldn't?
- compostor42 11y ago
  The idea of blue-collar workers getting well-paid is very upsetting to some white-collar workers. They feel their social status threatened by people they deem beneath them.
  - lambdapie 11y ago
    I don't think that's the primary issue people have with high wage union labor, and your comment shows an unwillingness to engage constructively with political opponents. Unions are a monopoly, and when prices (in this case, of labor) are high due to a monopoly, we must ask whether the government should allow, forbid, or encourage this monopoly. When wages are very low, I think most people are not opposed to unions forcing wages above market rates (I am, but that's another issue). But when wages are high, or driven far above what people think market wages would be, people question the moral basis for protecting workers from competition. So the issue isn't that blue collar workers are "beneath" white collar workers, but that (at least to an industry outsider) it's not clear what special skills these workers have that would command high market wages.
    - ConroyBumpus 11y ago
      My issue with union shops is that getting things done is unnecessarily difficult. Case in point: sitting around for half a day waiting for the unionized cable puller to sling a single cat5 connection between two racks on the same row in a data center. Then the guy doesn't show up until after 1600, so he forces overtime.

SixSigma 11y ago

The real automation in the Rotterdam port is the Automatic Guided Vehicle system that moves containers between the loading cranes and a container re-arranging system that optimizes the placement of containers dock side. The loading cranes themselves are only remote control on the deep water side, and this is for health and safety reasons. The cranes for the trains and barges for the hinterland are sit-in cabs. SOURCE: I did a tour of the terminal a few weeks ago as part of the 6th International Conference on Computational Logistics [1] A similar AVG system is in place at Hamburg as seen in this video [2] [1] http://realtimelogistics.info/iccl/ http://realtimelogistics.info/iccl/ [2] https://www.youtube.com/watch?v=3B3PzgYUK_0 https://www.youtube.com/watch?v=3B3PzgYUK_0

DenisM 11y ago
This video suggests the cranes are automated as well: https://www.youtube.com/watch?v=WxXZQ7emHC0 https://www.youtube.com/watch?v=WxXZQ7emHC0

Animats 11y ago

TraPac and Middle Harbor in the Port of Los Angeles are automating. Oakland isn't considered big enough to need it.[1] for an overview of automation in Los Angeles ports, see [2]. The Port of Los Angeles isn't a port operator; it's a landlord. Thirteen private companies have their own terminals, docks and equipment, and two of them are automating now. To automate a terminal, you have to take it out of service and build a new one. The above references may be too specialized for general readers. Here's a "gee whiz" video of Rotterdam port operations.[3] And a tutorial with animations, from the port of Busan, S. Korea.[4] Note that the article is partly an ad for FlexPort. Automated ports have less need for FlexPort's service. Most automated ports have buffers, big stacks of containers. Containers come off ships and are transferred to container stacks nearby. On the other side of the stacks, trucks approach and have containers loaded onto them. The terminal handles container sorting. Oakland loads many containers directly from ships to trucks towing container trailers, without buffering. Matching up trucks and containers is a problem. FlexPort's concept is "Uber for containers" (I know, but here it's real.) Any truck, any container, any destination within some distance limit. If Oakland had automated container stack buffers, FlexPort wouldn't be needed. [1] [http://www.joc.com/port-news/port-productivity/us-ports-weigh-value-terminal-automation-investment_20141002.html http://www.joc.com/port-news/port-productivity/us-ports-weig...] [2] https://www.portoflosangeles.org/Board/2014/April/042014_item5_Transmittal_3.pdf https://www.portoflosangeles.org/Board/2014/April/042014_ite... [3] https://www.youtube.com/watch?v=WxXZQ7emHC0 https://www.youtube.com/watch?v=WxXZQ7emHC0 [4] https://www.youtube.com/watch?v=2t2cqiQl0BE https://www.youtube.com/watch?v=2t2cqiQl0BE

fraserharris 11y ago
FlexPort's customer is the business paying for the shipping (freight forwarder & customs broker), not the port. Their interest in port automation* is that it makes for interesting content marketing and it may benefit their end-customer someday. * from what I can tell

Maarten88 11y ago

Rotterdam also has unions to deal with. A colleague of mine, who worked on the software for the final stage (some 15 years back) moving containers onto trucks, told a story where, after union negotiations, they implemented a system where forklift truck drivers could choose between three different containers to pick next instead of giving them just one instruction, for the single, most optimal, one.

macmac 11y ago
While the unions are arguably stronger on the US East Cost the International Longshore and Warehouse Union at the Oakland terminal certainly isn't a pushover.
cosmolev 11y ago
What is the logic behind it? Having an opportunity to select is more entertaining to the driver?
- chime 11y ago
  I am in the middle of writing directed pick/put module for a warehouse. I just had a meeting last week on the exact same topic. My software gives the user the most optimal pick list and the warehouse manager wanted me to change it so his team could pick from alternate bins. His justifications: 1. My software doesn't know that there is a cart/ladder/forklift blocking the most optimal bin while the bin next to it is easily accessible. 2. My software cannot tell if QC is in middle of sampling or if Inventory team is in middle of cycle counting that exact bin. 3. My software doesn't factor in changes in real-world that have not yet been inputted into the computer system. So if there is spillage, quantity, or quarantine issues that haven't been scanned into the system yet, my software calculates the optimal bin incorrectly. It's not that code cannot calculate what optimal means given a set of requirements. It's just that the end-user might be in a better position to decide if it really is optimal or not.

kalkin 11y ago

This article tells you basically nothing if you're not already convinced that more automation is always better, or that while "technological disruption... almost inevitably leads to displacement of some segments of the workforce. In the long run, however, technology ends up creating better jobs and expanded opportunities across broad spectrums of the economy." The calculation that "proves" Rotterdam is more efficient is based on the unargued and generally false assumption that both ports are operating at full capacity - the author just divides TEUs by number of cranes. TEUs handled per year in Oakland only just matched their 2006 peak after falling drastically during the Great Recession: http://www.portofoakland.com/maritime/containerstats.aspx http://www.portofoakland.com/maritime/containerstats.aspx And Oakland's main capacity limitation actually seems to be congestion getting containers out of the Port: http://www.wsj.com/articles/oakland-port-proposes-trucking-fees-to-fight-congestion-1440103747 http://www.wsj.com/articles/oakland-port-proposes-trucking-f...

tajen 11y ago
I was also expecting hard figures about how many containers and shifts have been lost in accidents compared to an automated port.

twic 11y ago

How is there an article about the costs of using containers without automated infrastructure on Hacker News and nobody's made a Docker/Kubernetes joke yet?

lgas 11y ago
Why didn't you do it?

vool 11y ago

''But in many ways, the Port of Oakland, and most American ports in general, are some of the most technologically antiquated in the world.'' This BLDGBLOG post from earlier in the year details to automation on the on the East coast. http://bldgblog.blogspot.ie/2015/04/infrastructure-as-processional-space.html http://bldgblog.blogspot.ie/2015/04/infrastructure-as-proces...

Someone 11y ago

Earlier discussion (different article, but same subject): https://news.ycombinator.com/item?id=10306204 https://news.ycombinator.com/item?id=10306204 (Other article at http://priceonomics.com/why-arent-americas-shipping-ports-automated/ http://priceonomics.com/why-arent-americas-shipping-ports-au...)

joecomotion 11y ago

IIRC, Rotterdam had a massive govt. subsidy for going fully-automated, which kind of helps. I spent 10 years between 1997 and 2009 (took a break for grad school) working on semi-automated tracking systems for ports up and down the West Coast (and other places), including one in Oakland. The recent surge in interest in intermodal container ports sure brings back memories! My company developed niche systems for tracking container location within the terminal by instrumenting and tracking any crane/truck/etc. that could move a container. Even though longshoremen were still operating the equipment, it helped to have a real-time (and historical) picture of what happened, because planning systems were now less subject to clerical mistakes. Here's their (sort of terrible) website: http://www.nowsol.com/ http://www.nowsol.com/ These guys pretty much own the market for the terminal-planning systems we fed: http://navis.com/ http://navis.com/

remarkEon 11y ago

From the Wikipedia page, it looks like Rotterdam had 36,315 vessel arrivals in 2008 while Oakland had only 1,928. Is this comparison really apples to apples? They point out the difference in tonnage, but I'm just wondering if there isn't a better comparison to be made. https://en.wikipedia.org/wiki/Port_of_Rotterdam https://en.wikipedia.org/wiki/Port_of_Rotterdam https://en.wikipedia.org/wiki/Port_of_Oakland https://en.wikipedia.org/wiki/Port_of_Oakland

jpatokal 11y ago
As noted in the article, Oakland is containers only while Rotterdam handles other types of cargo as well, and their comparison is based purely on containers. However, all things being equal, it should be more efficient to unload a few ships with lots of cargo compared to lots of ships with little cargo. That said, Oakland is somewhat weird choice, it's the world's 62nd busiest (by TEU) when Rotterdam is 11th. LA is the busiest in the US at 18th, with Long Beach (also greater LA) not far behind. https://en.wikipedia.org/wiki/List_of_world%27s_busiest_container_ports https://en.wikipedia.org/wiki/List_of_world%27s_busiest_cont...

exar0815 11y ago

This basically only shows again what the tiny Netherlands can achieve. I dont think there is a single country that threatened by a force of Nature. But instead of giving up, you basically become the best Worldwide in construction in, on and around the water. Basically, if you have any problems with water at any Kind of level, just ASK the dutch.