The competition is surging ahead. A few past public product failures are casting doubts on the ability of the company. The Boss feels under pressure to show the competition up, and he is pressing his product development team to deliver something that will blow the completion away.
A new product is planned that will show the world how powerful and important The Boss and his company are. The company had built a similar product which worked great, but this new design has an important new feature and there are no analytical models to predict the performance. The designer is the best in the business, but he is nervous about the design. It goes beyond his experience, and there is no time to slowly move up the learning curve. The Boss wants it his way.
The original designer leaves the project part way through. Management keeps changing the spec, to add more features that will “impress the world.” The product is built, but when it is secretly tested, the test manager has to suspend the test or the handcrafted first article would have catastrophically failed. The Boss with the very strong personality is out of town and can’t be reached, and no one is brave enough to tell him of the failure. They hope it was a fluke.
Production moves ahead and the product is released to great fanfare. A huge party is held to celebrate the launch of this new product, the company’s flagship creation, the state of the art, what will surely show the world how great their leadership is.
And what a magnificent disaster it is. In less than an hour of operation, the product destroys itself; lives are lost, all under the eyes of friends, families and competitors. The response of The Boss, who is still out of town: “Imprudence and negligence must have been the cause, and the guilty parties must be punished.”
Does this story sound familiar? Is it about the latest PC operating system? Is it about the last product your company wanted to ship, but you kept failing FCC certification and missed the market window? Is it about the first high speed serial link product your company tried to bring to market, without enough signal integrity experience to avoid the common landmines?
This is the story of the Vasa, a sailing ship launched on Aug. 10, 1628. The Boss was King Gustavus II Adolphus of Sweden. At the time of construction and testing, the King was leading his invading army into Poland, counting on his new battleship to help win the war.
Figure 1. A scale model of the Vasa with the original painted wooden sculptures, designed to show the world the might and power of the King of Sweden, in display at the Vasa Museum, Stockholm, Sweden.
Playing catch up
Sweden was behind the rest of the world’s navies and only able to build single-gun deck ships. Gustavus II wanted two decks of guns to project Sweden’s world power. But, the design of ships was an art, passed down from father to son. No designer in the shipyards of Skeppsgarden, nor in the entire Swedish Empire knew the numbers for dimension and ballast to ensure stability with the height and weight of two gun decks.
The original designer, Henrick Hyberstsson, had tried a conservative design and specified 24 pound guns for the lower deck and lighter weight, 12 pound guns for the upper deck. He died a year before construction finished. The King decided he wanted to project more power and ordered 24 pound guns for the upper deck as well. The new ship builders went along, since the King had signed off on the design.
As part of the stability testing of every ship, 30 sailors would run from side to side ten times to try to capsize the ship. When testing the Vasa, the ship master, Joran Matsson, had to halt the test after three cycles, for fear the ship would capsize. Admiral Klas Fleming, one of the most influential admirals in the Navy, witnessed the test. It was his responsibility to tell the King and halt the design, but he was too frightened of the King’s response and never mentioned the stability problem.
At 4 p.m. on Sunday afternoon, Aug. 10, 1628, the ship was officially launched with 90 sailors and their families—wives and children—aboard to celebrate the pride of the Swedish Navy. It was to be a short cruise around the bay and back again.
Using only four of her 10 sails, she pulled away from the pier. The Council of the Realm described the events in a letter to the King, who was still away waging war in Prussia:
“When the ship left the shelter of Tegelviken, a stronger wind entered the sails and she immediately began to heel over hard to the lee side. She righted herself slightly again until she approached Beckholmen, where she heeled right over and water gushed in through the gun ports until she slowly went to the bottom, under sail, pennants and all.”
The maiden voyage lasted for 1,300 meters, at the end of which 25 men, women and children lay entombed in the sunken ship.
No one responsible
An inquiry was held a month later. After hearing testimony from the captain, the sailors, the shipbuilder and the admiral, no one was ultimately punished for the disaster.
On the morning of April 24, 1961, 333 years after she sank, the Vasa was raised from the sea floor and now is preserved, intact, in the Vasa Museum in Stockholm.
Figure 2. The Vasa, brought up from the bay after 333 years and restored to nearly her original condition, on display in the Vasa museum in Stockholm, Sweden.
Why did the Vasa sink? Ultimately, it was a combination of being top heavy with too much weight in the masts and the two decks of guns, with not enough ballast. Even the ballast was not designed well. It was composed of round, river rocks which would roll with the ship, adding positive feedback to induce the ship to roll even more.
Figure 3. A model of the cross section of the Vasa showing the two gun decks and the rock ballast in the keel. Too much weight above the waterline and not enough ballast contributed to her sinking.
The legacy of the Vasa, suggests a few pointers for advanced product development that might still apply, almost 400 years later:
- Express your concerns when management changes the specs in the middle of the product design.
- If you are pushing the envelope of performance, there is no substitute to having an analytical model to accurately predict performance before you commit to hardware.
- When you do have first article and perform test and measurements, use the data to verify how well it matches the predictions and when it doesn’t, use the data to “hack into” the design to determine its limitations.
- Never hesitate voicing your concerns to management. The last thing they want is a surprise.
