April 13th, 2015

Paperless Office, container shipping and trucks (with some self-driving)

Yesterday got me thinking about getting rid of paper again. I was talking to my friend K. on the phone and telling R.'s (husband) story about how at his office, they were working down a stack of printer paper bought decades ago, since printing occurs so infrequently anymore. Whenever he does go use the 3-in-1, he has to wake it up, and he uses the scanning feature about twice as often as anything else. K. said her company (in Seattle) had completely gotten ride of printers, including the business division that used to do job printing as a service to other companies. They still had some fax machines (perhaps I misremembered this part?) and they had a special oversize printer with special paper for the designers.

At Applebee's for lunch, they have these tablets on the tables ("Presto" labeling) where you can order (sort of) and pay. If you are quick, or you tell your server you intend to, you can pay and have the receipt sent to your email and generate no paper to take home with you. I like that! I also don't mind the tip calculator. I know that some people believe these technologies ratchet up tipping, but I got into the habit of really high tips when the kids were little to apologize for our obnoxiousness. They aren't so little and are actually quiet and non-mess generating, so I don't need to do that anymore and the tip calculator helps me leave an appropriate sized tip.

I had commented to K. on the phone that Seattle tends to run way ahead of the Boston area in terms of adoption, and I used as an example that you just don't get paper receipts. But it turns out K. makes a point of collecting all of her paper receipts, because she saves them for taxes. Because there is no income tax in Washington State, the sales tax is deductible, and while there is a no-doc amount that you can claim, if you keep the backing paperwork and you spend more, you can deduct more against your Federal Income Taxes. Because K. is Amazing, she does exactly that, and it is a meaningful amount of money for her family.

The point I was making overall was that we used to talk about the paperless office a lot Back In the Day. It is now truly here, and yet we talk about it very little. And that is unfortunate, because I believe there is a lesson about technology adoption to be learned from the paperless effort. (OK, there are many I am sure!) Specifically, just like any other process automation, you can get some big wins in terms of reduction of labor usage and waste and so forth by automating _parts_ of the process. But the truly incredible wins come from when the output of each part can be input to the next part, when there is flow from one automation component to the next. Paper persisted because it was where we sent the output from office automation (and then sent it via interoffice mail to the next bit of office automation, or to be filed). Paper _proliferated_ in the office during the process of automation, because each new automation component freed up resources to take on more business and hence, more paper for those sections that were in between automated pieces.

There was a point in container shipping where at least some people really believed those boxes could be lifted off of ships and placed directly onto rail (and vice versa). A few docks were built with rail to the dock. At least according to _The Box_ that didn't work because it slowed the unloading process down too much, waiting for the train to move so the next slot to unload would open up. So instead, we use trucks and we have chronic air pollution problems in Long Beach and similar West Coast ports.

I had believed for years that we would fix this by finally getting rail out to the docks. Reading _The Box_ makes me think that that may never happen (again), partly because I don't understand, but when I don't understand part of reality and it still works that way and there is _that much money_ on the line, I'm inclined to defer to reality (I know! Shockingly humble of me. <-- sarcasm). I try to limit my arguing with reality to low dollar situations. And it further reminded me of a self-driving niche called "platooning", that has already seen real world testing but rarely comes up because it doesn't fall into the dream of your car driving your wasted ass home and then delivering your kids to school even if you are super hung over still.

We've been talking about _car_ platooning as a way to potentially get more traffic onto existing roadways (basically, don't let the human decide when to speed up, slow down or otherwise influence the space between vehicles -- turn that over to some automated process and run the jam as a platoon). There's even a wikipedia page for automobile platooning and none (that I can find) for truck platooning. That page talks about truck + car platooning. But truck platooning seems to be much further along.


All these technologies are originally tested on test tracks -- private roads. The Dutch stuff, however, has advanced to Rotterdam's Port, and when I first saw coverage, I foolishly thought that was because the Port was a controlled place with a lot of space for driving trucks around. I no longer think so. I think this is all part of a long running plan to finish automating the shipping process. The payoff for automating out of existence truck drivers is much more immediate than automating out of existence human drivers (or part of their focus and attention). If you platoon two trucks, you saved yourself the cost of one driver, who you were paying. There is no payment to divert or eliminate in a typical consumer driver situation, so showing how you make money by replacing them is much more difficult. Further, platooned trucks that get out of the Port and make way for more trucks faster mean that the biggest ships in the world can load and unload faster and make still more money -- again, a calculable benefit, in a way that is difficult to do with commuters on highways (and further, the commuters on highways could probably get the same effect of more people moving along an existing way at a particular time through better usage of shared transport, and that's probably a lot cheaper to deploy, expensive as it tends to be).

I'll be going on and on (and on and on) about the importance of seamless flow to truly benefit from a given technology wave.

ETA: Last month there was a 2 truck platoon on road test for Automotive Week in the Netherlands.


The US efforts on truck platooning are oriented towards something truck drivers do already: drive at the correct distance from each other to reduce air resistance/improve air flow over the trucks. Basically, Truck Peloton.


Platooning with cooperating vehicles is much easier than rolling out something in the consumer car fleet, because you can make assumptions (such as that the other truck you are platooning with has the exact same equipment that you do) that you otherwise won't be able to make for a long time if ever.

Port of LA _does_ have some on-dock rail facilities:


Port of Long Beach is developing on-dock rail facilities:


Tacoma (I realize at this point this is small potatoes -- little joke there -- compared to the above mentioned ports) has some on dock rail:


Weirdly awesome slide deck from ITS about the LA freight corridor, and the progression of autonomous features.


Tax season!

Yeah, I know. Everyone reading this is going, but I filed mine in January. Well, good for you! You are a wonderful, hard working person and I'm glad you will have gotten all the way through the process in such an expeditious manner (<-- completely serious).

I am not asking for any sympathy when I note the following.

I did finally start entering data from pieces of paper and downloading electronic data into TurboTax today (my husband got around to this last Friday). Today, _after_ I got mostly through what I am able to do (my husband has the health insurance info, for example, and I am a little unclear on whether all the paperwork for the house property tax is in the folder), I received another envelope in the mail marked Important Tax Information Inside. It had a paper version of one of several incorrect forms I've received for a thing that ordinarily is quite well behaved and generates exactly one form that never needs to be changed, and gets it to me well before April 15. Fortunately, I _think_ I found the right number in one of the messages I received electronically. OTOH, I also got yet another electronic message from those good people this afternoon.

It doesn't really matter. I'm not gonna get final data from two other sources until late September (<-- wild optimism here) anyway, so I'll be filing for an extension with the rest of us schmucks. And it looks like the WAG (<-- wild ass guess) I put together a while back so I could make payments has largely turned out to be accurate. Hoodathunk?

Delphi's cross country tour, stopping at SXSW coverage (linkage)


Includes a picture of the car which looks surprisingly normal (no antennas sprouting from the roof to a conspicuous degree). Delphi claims 99% self-drive. Human took over during a construction zone. Things Delphi is willing to admit happened that they intend to learn from:

"Passing or being passed by tractor-trailers, the car wanted to move over a bit farther than necessary (perhaps just as humans are wont to do. Conversely, it didn’t want to move to the left lane to give space to an emergency vehicle on the shoulder, something that’s becoming law in many states."


Piece on Kathy Winter, VP of Software Services and the trip. Funny error in the writing:

"“The biggest thing we learned from the trip was that the performance was better than expected,” said Kathy Winter who road in the car for some of the Delphi cross-country road trip to New York." Road should be rode. There are numerous errors of this sort in the writeup: very should be every, feed should be fed.

Delphi addressed a typical sensor problem of crap (snow, leaves, etc.) obscuring sensors in this way:

"The radar was not affected by dirt or weather because it is placed behind panels in the car. Clear lenses were cut into the bumpers for the Lidar. Lidar was also positioned below the back window." They did not clean the car during the trip, so the crap that accumulated on the car during the trip did not interfere with sensor functioning. Impressive!

They were able to avoid stuff on top of the car by replacing the spare tire in the trunk with what otherwise might have gone on top. But the backseat was available for and used by passengers.

"The trip was primarily on the highway because urban areas have to mapped in order for the vehicle to drive correctly." It's unclear whether this means, they need GPS mapping to work or something more detailed (that is, is this car driving in reality or is this car driving on a map? The google car, for example, seems to predominantly drive on a map. None of this is dependent on infrastructural support such as roadside beacons, which was the 90s approach).


Delphi's press release.


The blog, with pictures.

Quartz piece on cash in a mobile age


The chart is amazing: lots of cash still in use, notably for small payments, which payment cards are not great for. Then again, there are also a lot of people who never switched to payment cards; maybe ATMs can become the place where they are accustomed to paying their bills. Some QR code stuff for transferring money (put money into an ATM, get a QR code, send the QR code to someone else, they bring it to an ATM and get the cash. I think.).