Delivery drones of yesteryear: Rohrpost

Our modern world has given us many technological marvels and solved numerous practical problems. A visitor from 100 or 200 years ago would find many things to rejoice over. Airplanes have cut transportation times across the globe. Pharmaceuticals have erased many diseases. And computers have transformed work and entertainment in ways hardly comprehensible for someone from the 19th century.

Then there are the things that have not seen quite the same development. One is the delivery of goods. While information exchange has been transformed by telegraphs, telephones, fax machines and computers; the delivery of goods is still quite old-fashioned. When I order something online it will take quite a lot of time before a human shows up at my door with a parcel. Admittedly this human will come by delivery van rather than by wheelbarrow, but apart from that there is not much difference from the 19th century.

Interestingly enough, someone from the 19th century would scarcely believe how primitive goods deliveries in the 21st century would be. Parcel transportation was something that was in flux in the latter part of the 19th century. New technologies were being developed which promised to revolutionize the transportation sector.

Tube life

One of these revolutionizing new technologies was the pneumatic tube. In theory it is a very simple construction. A cylindrical container is placed in a sealed tube, air pressure is applied and the container will travel down the tube to its destination, which can be several kilometers away if the air pressure is high enough.

This was well within the technological capabilities of the second half of the 19th century and during this time there was great interest in this new technology. Usually the systems were designed to transport letters and very small parcels. They were a sort of speedier postal service, excellently summed up in the German word Rohrpost (literally “tube post”). Since German words are underutilized in English and since this article is specifically about pneumatic tubes as a substitute for postal services, the term Rohrpost will be used henceforth.

Most major cities in the developed world acquired some sort of Rohrpost in the late 19th or early 20th century: London in 1863, Berlin in 1865, Paris in 1868, Prague in 1889 and New York in 1897.

Contrary to what one might believe, these systems were not in competition with the relatively new telegraph systems. Rather they complemented each other and the Rohrpost systems grew on the back of the telegraph systems. While the telegraph was unbeatable on long distances the Rohrpost complemented it by significantly speeding up the manual delivery of the telegraph messages.

In practice, a message might arrive at the central telegraph office from someplace far away. There it would be written down on paper and this paper was transmitted through the Rohrpost to a local post office where a runner would be sent out to immediately deliver the message to its recipient. This system worked well until real competition materialized in the form of telephones installed in private homes.

Reality, always the reality

Rohrpost might have been able to compete with telephones if there had just been as many Rohrpost connections as there were telephone connections. But while telephones were installed liberally in private homes and offices, Rohrpost never progressed beyond post offices and some larger offices (mostly governmental).

The reason for this was economic. Compared to an electrical wire, required for a telephone connection, a Rohrpost tube is very expensive. There were also technical problems. One problem that was never satisfactorily solved in Rohrpost systems was switching, moving a container from one line to another. This meant that Rohrpost always consisted of one single line and at each end a human had to manually transfer the package to the next line. In theory this was not much different from early telephone systems. But while a telephone switchboard operator could handle hundreds of individual telephone connections a Rohrpost operator could handle only a handful. The picture at the top of this article is from a late 19th century French Rohrpost office and gives an inkling of the size and complexity of early Rohrpost systems.

As soon as telephones had established themselves Rohrpost found itself in a very difficult environment. Most Rohrpost systems saw their maximum usage in the 1920s and 1930s. From then on it was downhill. In Germany the war did its part by destroying much of the Rohrpost infrastructure, especially in Berlin, where the Western portion of the system was never restarted again, despite being one of the world’s most extensive before the war.

Post-war, hardly anyone took Rohrpost seriously. Major Rohrpost systems were quietly disabled for example in New York in 1953 and Vienna in 1956. Some Rohrpost staggered on, most notably the one in Prague which was in active use until 2002 when flooding disabled the system and repairs turned out to be too difficult and costly for it to be worthwhile.

The Prague Rohrpost was slightly modernized over the years but it was still obsolete as the time of its disabling. The most modern public Rohrpost probably was in Paris. Paris had an extensive and relatively well-used system and during the 1960s and 1970s some effort was made to modernize it. Some electronics were installed, details on this are scant but it probably was some type of electronic signal system. Also, some of the old steel pipes were replaced with plastic substitutes. Plastic has less friction than steel which makes the ride faster and smoother. Plastic is also not susceptible to corrosion which had been a problem with the old steel pipes. All efforts were in vain, however, the Poste pneumatique de Paris was closed down in 1984.

The main problem

Almost all of the Rohrpost systems of old were designed to transport messages. Some systems did have the ambition to move around parcels and other larger stuff, most notably the London Pneumatic Despatch Company which was not so much a Rohrpost system as a pneumatically powered railway. Despite building a 500 meter long operational track it never progressed beyond the experimental stage.

In the 19th century, when most Rohrpost systems were created, there was still a great demand for the transportation of messages. Less so for the transportation of goods. Keeping tubes small not only saved money on construction but made running the system significantly cheaper. There were technical reasons as well, all early pneumatic systems used steam engines for power, large, noisy and ineffective beasts. Everything that could keep power needs down was beneficial.

Most of the old Rohrpost systems therefore used tubes of a rather limited diameter. Both the Paris and the Prague systems had an inner diameter of 65 mm. The Prague system was quite strong and could transport up to 3 kg in a single container. But since the containers were 48 mm in diameter and 200 mm in length it was still very limited in the types of goods it could transport.

Old school 65 mm Rohrpost container from the Prague system

If the Rohrpost had been used more widely for goods transportation it might have been able to survive the onslaught of telephones and computers. Tubes and containers would have had to be bigger, at least somewhat bigger, to be able to hold more than just letters and documents. Such parcel delivery would have given Rohrpost a market which telephones and computers could not threaten.

A bigger problem is the network. To be truly useful a Rohrpost system should reach each and every domicile. This was prohibitively expensive in the 19th century and it still is prohibitively expensive. A hundred years ago this was not that much of a problem since labor was plentiful and cheap. Rohrpost’s edge over dedicated couriers was more about speed than price. But speeding up the last kilometer of delivery by building Rohrpost into every home simply did not make any sense at all.

Labor is more expensive today. And investment capital is more plentiful. But still not plentiful enough to pay for an extensive network of Rohrpost, at least not to the general public. However, there might be limited areas where the demand for quick and inexpensive transportation is high enough to offset the massive investment. Interestingly this is precisely what has happened.

Scraps of the future

While public Rohrpost never was that much of a hit and is now completely defunct there is a less known but more successful private variant. Even in the heyday of Rohrpost there were always minor systems installed in local areas, for example in offices or department stores, serving as a quicker way to send messages and small items than running around with them. These private systems had the same problems with cost and depreciating technology as the public Rohrpost and most seem to have been retired.

Today one type of institution towers above all others when it comes to pneumatic transportation: hospitals. This seems to be a modern phenomenon, I have not found any instances of hospitals using Rohrpost systems during the 19th or early 20th centuries. The reason for this is probably that hospitals do not use Rohrpost in the traditional way, to transmit documents and messages, instead they use it in a modern way, to transport small packages.

There is a reason that hospitals were late into the Rohrpost game. A big part of hospitals’ transportation needs consist of sensitive goods; laboratory samples, including very sensitive blood tests, are among the most common items in hospital Rohrpost systems. A traditional Rohrpost system will have constant pressure in the tubes, hurling away containers with a constant force and delivering them at the destination with an audible thud. This thud would have ruined anything delicate being transported.

Hospitals did not start using Rohrpost until the 1960s and 1970s (earlier examples might exist, but I have not found them). Ironically at the same time as the last public Rohrpost systems were being dismantled. The reason for this was technology advancement. While old Rohrpost systems used centralized engines to generate air pressure, new systems use small blowers dispersed in the entire system. Combined with better electronics this gives the ability to control individual containers with great precision, including controlled acceleration and soft stops.

Data on hospital tube systems is scant but most major hospitals in the developed world seem to have some sort of them. Stanford Hospital boasts about having one of the largest ones around. Its total tube length of 6 km is not impressive compared to the Poste pneumatique de Paris (427 km at its greatest) but it has 124 stations which is almost as many as the 130 in Paris. And at 2.5 million packets a year it has about the same volume as the Paris system had in its twilight years in the 1970s.

The most impressive fact about the Stanford system is that it is fully automated. The human role is limited to placing the container in the terminal and dialing in the destination. Everything else is done by technology. The chief engineer also claims that they have never had a stuck container and very few malfunctions overall.

A capsule in time

Hospitals are a niche application for a technology that once was destined for the mass market. But since hospitals were never involved with the original Rohrpost it might be more correct to speak of them as an entirely new iteration of an old idea. This is not a bad thing. The Rohrposts of the 19th century were founded on the premise of message transport, which would have been of limited use today, even if the systems had been modernized and remained in use.

In contrast, hospital systems were designed from the outset to move parcels around. For this reason they generally have wider dimensions and larger containers (the de facto standard seems to be 160 mm tubes carrying containers 15 x 30 cm in size). They are also made to be cost-effective which has led to them being highly automated and decentralized.

Modern 150 mm container from Aerocom

Hospital systems could therefore be used as a template for a modern city-wide Rohrpost system, a Rohrpost system made for parcel delivery rather than letter delivery. At the very least the demand for hospital systems means that technology development is ongoing. Multiple companies design and manufacture pneumatic tube systems, which mean that they will not be confined to museums for the foreseeable future.

Back-of-the-envelope

One piece of information that is very difficult to obtain about modern pneumatic tube systems is its cost. These are all bespoke systems and there are no official price lists. I even tried to email a few manufacturers of pneumatic tube systems for the health care sector but none deigned to answer. Finally, after much googling, European public health came to the rescue. Danderyd’s hospital, one of the main hospitals of Stockholm, for some reason did not have a Rohrpost system until one was installed in 2014. Being a public entity it supplied full details of its investment as described in this magazine article (in Swedish): Danderyd’s hospital invested 13 million US dollars. For this money they got 5 km of tubes and 110 stations.

110 stations for 5 km of tubes correspond to about 45 meters of tube per station. Presumably that would be enough tubing per station to connect a suburban residential area with one station per residential unit. The cost per house would then be 119 000 dollars.

This is quite a lot of money. In fact it is so expensive that it makes modern Rohrpost clearly unfeasible and by itself explains why it should not be considered seriously. But, being a bit unserious, just how unfeasible is it, really?

The headline investment cost is huge. But maybe it is inflated. After all, the numbers are from a hospital system, with the quality and cost that usually comes with everything health care. It was also a bespoke, retrofitted system. A significant part of the cost most probably was designing and modeling the system. Building a standardized system from scratch might turn out significantly cheaper.

Then there is the cost of the competition: delivery drones. According to an article in Business Insider Amazon projects its Prime Air drone deliveries will cost 63 dollar per delivery. Compared to this, Rohrpost is suddenly competitive again. With a discount rate of 4% the capital cost of the Rohrpost above would be 4764 dollars per year. This is the same as 75 Prime Air deliveries annually. There are of course operating costs as well, but the comparison is suddenly not as skewed as before. Especially if you consider that the average Rohrpost system has higher capacity and better reliability than delivery drones can ever hope to achieve.

The bottom line

Even state-of-the-art Rohrpost systems will have an uphill struggle to compete with airborne delivery drones. But not an impossible one, mostly due to the fact that neither technology looks very competitive compared to old-fashioned postal services. Both Rohrpost and delivery drones require great emphasis on delivery speed for them to be in the game at all. On this, Rohrpost has an edge over delivery drones. It also has an edge when it comes to reliability, safety and noise pollution.

Rohrpost has one almost insurmountable obstacle in its investment cost. Its operating costs are low but the investment cost is astronomical. Delivery drones are the exact opposite. It does not cost Amazon very much to start up an experimental delivery drone program, even if each delivery costs a small fortune. When the experiment is done it can be winded up and the losses will still be manageable.

An experimental Rohrpost system connecting a thousand homes would most probably cost hundreds of millions of dollars. Even if the Rohrpost turned out to be very useful it would still be a waste of money since such an experiment could not be reproduced, simply because there would not be enough money around to do it again on a larger scale.

The only way a Rohrpost experiment could be successful is if it proved that Rohrpost was less expensive and more useful than expected. The best chance for such an experiment to take place would probably be in a new up-market residential area that just happens to lie next to an Amazon warehouse. As far as I know no such experiments are under consideration and, frankly, they are not very likely to be either. 

It is considerably more likely that drone deliveries continue their incremental march towards success or failure. Rohrpost will remain a peculiarity whose future is always behind it.