Friday, 18 October 2024

Every Jetliner Ever Made: Bombardier

Let's have a look at Every Jetliner Ever Made. In alphabetical order, but with Airbus and Boeing at the end otherwise it would be boring. This week, Bombardier Aviation of Canada. Bomb-bar-deer, or Bomb-bar-dee-yay?

Apparently the latter. It was founded by Joseph-Armand Bomb-bar-dee-yay of Quebec, which explains why the company has such a warlike name. In the 1930s, initially specialising in snowmobiles, but it went on to be one of those classic 1960s conglomerates that made everything, at various times selling snowmobiles, business jets, buses, trains, and, yes, jetliners.

During its time as a jetliner company Bombardier only produced two basic designs, but one of them was very popular, nay era-defining.

A Bombardier snowmobile, photographed by Jim Peaco of the US National Parks Service

BOMBARDIER (Canada)

Bombardier's speciality was regional jets. Regional jets were so hot in the 1990s and early 2000s. Aviation fuel was cheap and people wanted to go on holiday. What is a regional jet?

There's a fundamental difference between the commercial aviation market in the United States and Britain. If I want to fly to Athens from the UK I can take the train to Heathrow and then fly directly to Athens. But what if I lived in Cleveland? I don't, but imagine if I did. I would be a completely different person, but humour me. I would be hard and mean. Imagine if I lived in Cleveland and I wanted to go to Athens. To get away from Cleveland.

Cleveland doesn't have flights to Athens. It doesn't. I've looked. It does however have flights to New York. If I could just get to New York first, I would be able to go to Athens.

But how to get to New York? In theory I could drive, but it would take ten hours. And the United States doesn't do trains. So I fly. And I would not be alone, because Cleveland and its surrounding area has a population of over three million people, and Cleveland is just one city.

Enough of Cleveland. Suffice it to say that in the United States the thought of flying from one city to another to catch a second flight somewhere else is not unusual.

Cleveland, 1986. Invented by Moses Cleaveland, who left after spending three months there and never returned.

It's called the hub-and-spoke model. The smaller airports are spokes, the larger airports are hubs. As a consequence the US has a substantial market for super-small jetliners that can cover short distances without using too much fuel. They're called regional jets, sometimes feederliners, and that's where Bombardier comes in.

And Fokker, and De Havilland Canada, and Embraer. All of those companies specialise or specialised in small regional aircraft that could fly between cities in the United States. The hub-and-spoke model also operates in South America and Asia, but I have the impression that Asian carriers prefer larger, full-sized airlines.

Up until the 1990s regional routes were usually covered by turboprops, such as the De Havilland Canada Dash 8 pictured above, but turboprops have some disadvantages. Limited cruising altitude, slow speed, and noisy engines among them. Furthermore there was a perception in the 1990s that turboprops were old-fashioned. When questioned, passengers frequently expressed a preference for jets, and in 1993 Comair, a subsidiary of Delta Airlines, took the bold step of ordering a batch of regional jets from Bombardier, which opened the floodgates for other airlines to do the same thing.

Sub-100-seat regional jets went on to become a huge and hyper-competitive market in the 1990s, largely killing off turboprops. Dornier struggled to introduced a regional jet, and although Fokker had a head start with the F28 Fellowship, which had pioneered the concept, the development cost of their own modern regional jet left them heavily in debt.

A Fokker 100 regional jet. The regional jet market was hyper-competitive, which unfortunately resulted in Fokker going bankrupt in 1996.

Britain is too small for the hub-and-spoke model to make any sense, and continental Europe has a high-speed rail network that's generally cheaper than an airline ticket, but a few European airlines do use regional jets. Scandinavian, for example, has a fleet of Bombardier CRJs, as does Lufthansa. Regional jets make a certain amount of sense on the edges of Europe, especially the Nordic area, where the population is widely spread out.

After the failure of Dornier and Fokker, and the withdrawal of Saab and BAE Systems from the regional market, the two remaining players in the immediate pre-COVID period were Bombardier of Canada and Embraer of Brazil. Following behind them were the Franco-Italian ATR and De Havilland Canada, but they make turboprops so for the purposes of this document I'm going to pretend they don't exist. I'll cover Embraer in a separate document, but suffice it to say that they were Bombardier's arch-rivals during the 1990s and 2000s.

A Bombardier Challenger business jet, in service with the Royal Canadian Air Force

Bombardier's regional jet programme grew from its experience with business jets. The company's first regional jet, the CRJ100, was launched in 1991. The company had gambled that the current batch of turboprop airliners were getting long in the tooth, and as mentioned earlier in the article the CRJ was given a huge boost by an order from Comair, who became one of the launch customers. Comair went on to order 110 CRJ100, becoming the type's main operator.

The CRJ100 had the same configuration as the Bombardier Challenger business jet, with two rear-mounted engines and a T-tail. A classic configuration for smaller airliners that had the benefit of keeping the engines away from runway debris while allowing for shorter landing gear. The biggest Bombardier business jet, the Challenger 650, could seat around twenty passengers, but the CRJ100 was much larger, with a passenger capacity of fifty people in a 2x2 arrangement, and a range of around 1,500-2,000 miles depending on the model.

It was conceptually similar to the Douglas DC-9 or BAC 1-11, but with a much smaller passenger load. Why CRJ? The programme was originally the Canadair Regional Jet, but Canadair was bought up by Bombardier in 1986. In fact De Havilland Canada was also bought up by Bombardier, but it was then sold again because Bombardier didn't want to make turboprops.

A Bombardier CRJ 200, courtesy contri of Japan (CC-SA 2.0)

Production of the CRJ100 was modest, amounting to only a couple of hundred sales, including 110 to Comair and 35 to Lufthansa of Germany. However in 1996 Bombardier launched the CRJ200, which had more efficient engines but was otherwise much the same. It sold like hot cakes, and of the 1,000 or so CRJs ever made the production ratio was about 7:3 CRJ200s to CRJ100s. They were mostly bought by airlines in the US, predominantly Northwest Airlines and Skywest. Production continued until 2006.

I'm unfamiliar with the regional jet market, but the late 1990s was apparently its heyday. I have the impression that the twenty-first century as seen from 1999 consisted of regional jets ferrying passengers to New York and Los Angeles forever.

A CRJ 701, with an obviously stretched fuselage.

By the late 1990s the CRJ100/200 was pushed aside by the CRJ700, which was launched in the late 1990s. It had a shorter range than the CRJ200 but carried around 50% more passengers, around 70-80 or so, with the same two-engines-at-the-back configuration. The launch customer was Brit Air of France, who began operations in February 2001. During the 2000s Bombardier also released a pair of stretched versions of the CRJ700, the CRJ900 (2003) and CRJ1000 (2007), which carried around 90 and 100 passengers respectively. The three of them also sold around 1,000 units, the bulk of production being CRJ900s and CRJ700s.

There was also a pair of oddities, the CRJ550 and 705, which were essentially CRJ700s and 900s with fewer seats. There was no real technical reason for this; they existed because some airlines in the United States were contractually forbidden from operating regional jets that carried more than a certain amount of passengers. 50 and 76, apparently. I have no idea why.

A CRJ 900. Notice the extra exit door.

The 100-seat CRJ1000 was the top of the range, but it wasn't particularly popular. For legal reasons regional operators in the United States were prohibited from using it, but conversely it was too small for most customers in Europe, who were gravitating towards uniform fleets of Boeing 737s and Airbus A320s. Development problems delayed its first paying flights until 2011, and production only continued until 2020. A year later the CRJ line was sold off to Mitsubishi of Japan, who continue to provide parts and support for CRJs, although the company has no plans to restart production.

Nonetheless the CRJ1000 had lit a fire within Bombardier. There was a general trend in the industry for regional jets to grow and grown in size, so Bombardier decided to jump up a tier and release a full-sized airliner. A miniature full-sized airliner.

This was launched in 2013 as the CS100. It had two engines mounted under the wings and a conventional tail, just like a Boeing 737 or Airbus A320.

A Bombardier CS300.

It had a passenger load of around 120 people, with a range of 3,500 miles, considerably more than the CRJ. It was joined in 2016 by the CS300, which carried around 150 people.

The CS100/300 was controversial. Until that point Bombardier - and Embraer, and all the other regional aircraft manufacturers - had avoided directly competing with Airbus and Boeing. And, technically, the CS100 did not step on the toes of those two manufacturers, although that didn't stop Boeing from taking legal action against Bombardier for selling CS100s to Delta Airlines at a keen price.

The Boeing 737 and Airbus A320 loom large in the world of aviation. They were both originally developed as small airliners that could carry a useful passenger load across the United States or Europe, but over time they have been stretched and re-engined to a point where they can almost cover transatlantic routes. They have greater range and more capacity than the four-engined jetliners of the early 1960s, and in fact Norwegian Airlines briefly flew 737s across the Atlantic from Ireland to the US east coast in the pre-COVID years. But conversely they can, at a pinch, still fly the shorter routes for which they were originally designed.

Bombardier reasoned that the 737 had grown too large for some carriers, so the CS100 was intended to slot in underneath it. Unfortunately, despite their best efforts, regional airlines in the United States ignored it, and sales have mostly been to European and African carriers, so in 2017 loss-making Bombardier sold the design to Airbus, who now make it as the Airbus A220-100 and A220-300. Boeing was like "frumple".

As of this writing the A220 is still a going concern, although its reception has been muted. The advent of COVID in 2020 did nothing to help. It has sold around three hundred units, with firm orders for five hundred more. That's nowhere near the amount of orders won by the Boeing 737 or Airbus 320 over the same period, but it's not bad. Perhaps in a world of uncertainty over infectious diseases and climate change there's room for a smaller, more flexible alternative.

A Bombardier Global Express business jet - the company essentially went around in a big circle, starting and ending with business jets.

And that's Bombardier. As mentioned earlier, the original CRJ range was sold off to Mitsubishi, and the CS100 was sold to Airbus, so the company no longer makes jetliners. It does however still make business jets. The largest, the Global 7500, has a passenger capacity of around 19, too small for this document, but notable for an extraordinary range of almost 8,000 miles, which competes favourably with most full-sized jetliners. From 1990 until 2021 Bombardier also made the famous Learjet, which is also not an airliner but looks fantastic, viz:

Tuesday, 1 October 2024

Using AI to Generate Content: Yay or Nay?

Over the last few years there has been a lot of noise about artificial intelligence. What is it? What does it do? Can it be used to make money?

The majority of AI solutions are language learning models. They digest a mass of human-generated written media, and in doing so they learn how to generate responses that mimic human intelligence. On the surface the end result feels like a cheat, an imitation, but the fact is that human intelligence itself is an imitation of human intelligence. We are all just clever mimics.

I mean, have you ever tried to have a conversation with a child, or anybody under the age of about fifteen? They don't understand anything. Nothing sticks. There's nothing inside them. The words just go into their stupid faces and straight out again. The fact is that young people are clever mimics of humanity, without the depth of thought that comes with experience. They can be trained to carry a rifle, but they are empty, empty vessels.

AI is a lot like that. AI is an empty vessel, a precocious child, somewhat akin to the population of Reddit. AI is horrifying, in a way, because the notion that we are just a mass of programmed responses to external stimuli implies that we are not divine at all. I'm digressing here.

The key thing is to train the model on a huge, diverse pool of good-quality written content that also contains genuine specialist information, so for the purposes of this blog post and potentially the entire future of this blog I decided to use UFOPaedia, a website that covers the long-running XCOM series of tactical wargames. It has been around for many years and contains a mass of fascinating information. I could have used Wikipedia, but I don't want to make a blog that just lists Pokemon.

My goal was to see if I could use AI to write an informative article about internet content generation for a general audience. If this works, I might be able to go on holiday for once. I can hit "go" and let the AI generate the rest of this blog while I go on holiday. At last, I will be able to go on holiday. Not just writing. Also holiday. Let's see how it goes. I begin, although at the back of my mind I worry that I might have chosen too small a dataset.

1. USING AI TO DRIVE TRAFFIC TO YOUR WEBSITE

It's pointless having the wittiest, most informative content on the internet if no-one ever finds out about it. You need to drive traffic to your website, and the best way to do this is to MANUFACTURE LASER CANNON.

A laser cannon

Of all the potential solutions, laser cannon are by far the best. Unlike almost every other piece of technology that exists in this world, a laser cannon does not require raw materials, only money. Laser cannon have a flat cost of $182,000 and a sale price of $211,000, making a paper profit of $29,000 per cannon, although in practice this doesn't include the maintenance cost of a workshop and the wages of the engineers required to build the cannon.

Nonetheless, with production running 24/7, each cannon generates a profit of $18,000 - and a single fully-staffed workshop can produce just over 107 of them a month, for a monthly profit of over $1.9m. That's a lot of money.

2. USING AI TO DRIVE READER RETENTION

Suppose your readers get bored, and decide to leave? You could in theory sate their lust for mental stimulation by giving them more content, but a more effective way of retaining readers is to MANUFACTURE LASER CANNON.

An XCOM engineer manufacturing a laser cannon

As mentioned previously laser cannon have no special material requirements. There is a minor setup cost involved in researching laser pistols, laser rifles, and heavy lasers, but each cannon only requires money and engineer hours.

The next-most-profitable technology is the FUSION BALL LAUNCHER, but each launcher requires one unit of alien alloys, which could otherwise be sold for a profit. When the sale price of the alien alloy is taken into account, laser cannon are more profitable than fusion ball launchers. Furthermore alien alloys do not exist. They are alien. They do not exist.

3. USING AI TO GENERATE NEW CONTENT

There is an upper limit to the amount of content a human being can generate, and technically I suppose the same is true of AI, but in practice the upper limit is much higher. After less than eighteen hours of non-stop writing a human writer begins to suffer mental anguish and fatigue, and beyond a certain point her writing turned into unreadable drivel interspersed with tearful cries for help.

In contrast AI never gets tired and does not require help or a working heater. It can produce fresh, inventive content forever. The most effective way to do this, of course, is to MANUFACTURE LASER CANNON.

XCOM operatives engaging a UFO scout ship with portable laser weapons (colourised)

In addition to the positives mentioned above, laser cannon also have another thing going for them. In the unfortunate event that one of XCOM's interceptors is shot down, the cannon are immediately useful. A pair of them can be immediately installed into a fresh Interceptor. In contrast, fusion ball launchers are useless without a payload of fusion balls, which are ruinously expensive.

As a weapon the laser cannon is inferior in every way save profitability to the plasma cannon, but a dual-laser Interceptor is still capable of downing all but the toughest of enemy UFOs, without the ammunition storage requirements of the Avalanche missile system.

4. USING AI TO EXPAND YOUR PORTFOLIO

Suppose you decide that a blog isn't enough. Perhaps you want to be a YouTube personality, or perhaps you aspire to appearing in print, or on television. The most effective way to do this is to MANUFACTURE LASER CANNON.

XCOM operatives celebrate after successfully downing a small scout with laser weapons (colourised)

Laser cannon have a third benefit. Once researched, they allow XCOM to research and manufacture laser-equipped tanks. The Tank/Laser has some limitations - its main weapon is less accurate than the standard Tank/Cannon, and by the time XCOM is able to field the Tank/Laser it will be facing aliens armed with heavy plasma rifles, which can destroy the tank in one shot - but on the positive side the Tank/Laser is the second-most-effective tank against Sectopods, on account of that creatures' vulnerability to laser fire.

Furthermore the tank's laser weapon has a roughly 5% chance of penetrating the inner walls of a downed UFO, less than half the chance of a heavy plasma round, but the tank has a battery capable of firing 255 shots versus the plasma rifle's 35, so each shot is essentially free.

~

So, in summary, AI is a powerful tool in the hands of the right author, such as myself. In the next instalment I shall see if I can come up with photography tips by pointing it at the Biker Banter section of Pistonheads.com. The most effective lens? Bacon sandwiches.