Bjorn’s Corner: The challenges of airliner development. Part 30. Serial Production

By Bjorn Fehrm, Henry Tam, and Andrew Telesca.

November 19, 2021, ©. Leeham News: Last week, we described how we finished the testing and the process to get our Type Certificate.

Now we look at the phase after Design and Production certification, the start of production, Figure 1. The upstart and ramp of production have many challenges. We will start the discussion with one that is often overseen, the cost of ramping production to full serial production rate.

Figure 1. Our program plan for the aircraft. Source: Leeham Co.

Serial Production and the learning curve

It’s now time to discuss the challenges involved when we start the serial production of the aircraft. Let’s first start with something many projects are not so clear about.

A common idea is that the main challenge of getting a new airliner to market is the development, both technically and financially. The next challenge is to get a smooth production going with deliveries from hundreds of suppliers. 

From a technical and organizational perspective, this is true. What is often overlooked in the setup of a project is how big a challenge the first part of the production will be financially. The culprit is the production learning curve.

In the first part of the production run, the aircraft will cost more to produce than a customer is prepared to pay. The typical aircraft production learning curve in Figure 2 is telling the story. It’s conceptually describing the production costs for our 19 seater.

Figure 2. A typical aircraft production cost learning curve, with net revenue from sales of aircraft superimposed. Source: Leeham Co.

The initial production doesn’t run as smoothly as our projections of mature costs promise “once we are at rate”. Mature pricing is often calculated around the 400 units produced production number (it depends on the size of our forecasted sales in the business plan). Our first units are close to three times more expensive to produce than this mature cost (if things run to plan, if not more).

Gradually the costs to produce decline as more a more kinks are eliminated in the assembly of the aircraft sub-parts and then the whole aircraft at our Final Assembly Line, FAL. The historic learning curve on the aircraft level has a 15% decline in production cost for every doubling of the produced number. So number 20 cost 15% less to produce than number 10, and number 40 15% less than 20. This is called the 85% learning curve.

Why this happens we will go through as we go through all the issues we face when starting up the production.

Revenue has an inverse curve

We have also drawn in what could be the typical revenue we could get for our sold aircraft. When our sales negotiate the first deals for our aircraft, which are often called Launch customers or sometimes Marquee customers, these want special pricing. 

They take a risk by betting on a new aircraft in the market. If the aircraft is not selling well they are trapped with an orphan where even the OEM might run into problems because of selling less than planned and he might cease to exist. Thus they demand special pricing and they get it (all OEMs prime the initial sales of a program with extraordinary deals).

It means we have an effect where the product costs more to produce and we get less revenue than we can get for say unit 100 or thereafter. This results in initial losses that can be substantial (area A in the graph). A rule of thumb is that the initial cost of production versus actual revenue can cause losses as large as the development costs. 

The initial part of the revenue curve is not a smooth growth. The reason is that special pricing for a customer applies to his whole delivery. Let’s say he orders 20 machines with an option that he later takes for 15 more. It means 35 aircraft of our initial production has low pricing, and the revenue curve only goes up as we enter an area with later deals that are not Marquee deals. 

Our losses on top of our development costs continue until we in this case have produced 100 aircraft (which can take three to five years) and the development costs and initial production costs will only get paid back as we reach area B and this has grown to cover initial costs.

Thus management and investors in the project need to understand that the development and initial production costs can run to double the development costs. The number of upstart aircraft manufacturers that have gone belly up in the initial production phase due to a liquidity crisis is a long list. In fact, it’s more common for new OEMs to fail during this phase than during development.

The very ambitious Personal jet company Eclipse Aerospace behind the Eclipse 500 failed for this very reason. The company did not fail on order volume, it failed on understanding its own costs. It had a too low selling price, decided early in the project when the initial productions cost overrun due to the learning curve were not correctly estimated.

The multitude of reasons for the learning curve and what OEMs have tried to do to improve on the curve’s ski-jump slope we discuss in the next article.

4 Comments on “Bjorn’s Corner: The challenges of airliner development. Part 30. Serial Production

  1. Often the initial product is over weight, poor reliability, does not meet payload and range numbers to 100% and cost too much to produce. Hence service engineering and production specialists spend wave after wave improving the quality and reducing price, sometimes the cost reduction introduces new problems so it normally takes some years and recalling retired experts to get it right and normally you have old products pumping in money to pay for it (like GE with the CF6-50/CF6-80C2, RR with RB211-535, Boeing with 737’s)

  2. Agree, Claes.

    We will drill deeper into the typical specification misses that the initial production aircraft has in the next Corners. But this is another reason for special pricing of the intial production run.

  3. One I am watching is the Celera Aircraft. Unusual concept and some wild top speed claims. (460 mph at 50k I believe)

    Seems to be a build it and they will come vs a confirmed market of orders.

    • Virtually none of its claims are achievable in practice …outside a computer
      Fully laminar flow when in practice its only partly achievable. Very dubious propellor efficacy claims
      A V12 piston engine running on diesel cycle but with jet fuel and turbocharged- intercooled to achieve its ‘computer numbers’ at 60,000 ft.

      Everyone else uses small turbo props for a good reason. The best small TP business class plane with many of those principles that are workable at 44k ft is the Piaggio P.180 which has a good sized interior for the weight of the plane. It hasnt been a runway success but sold steadily.
      The US and international market just loves jets

Leave a Reply

Your email address will not be published. Required fields are marked *