Korva is a scroll of the internet’s best ideas.
Design the ship and yard together
One owner for both
Historically, one party is paid to design a ship and then hands it off to a totally different party to build. When the same company owns the design and the shipyard, software engineers, shipbuilders, naval architects, and welders can eat lunch together in the yard, and optimizations appear that the old system was never set up to capture.
Designing from scratch
The Port Arthur site is greenfield, so the manufacturing plant for ships can be designed from scratch. It resembles hardware and software co-design on chips, where the compiler and the chip are designed to fit each other. A ship designed from first principles to be simple and built at scale also comes with training, work instructions, and processes made for it.
Halving the cost
The US builds ships at five to six times the cost of China. Saronic aims to cut the cost of a ship in half, so a $300 million ship would come in under $150 million, through the product, the process, and new shipyards. Wages go up along the way: welders and pipefitters get good wages, the same benefits as Silicon Valley software engineers, and equity in the company.
How did a greenfield yard build ships so fast?
Hyundai Heavy Industries began building its Ulsan yard in South Korea in 1972, and construction of its first ships started before the yard was finished. The first two supertankers were delivered in 1974 to the Greek shipowner Livanos. Designing the yard and the ships together let the company plan crane reach, block sizes, and workflow at the same time. The Ulsan yard has since grown into one of the world's largest.
Why does designing for the factory lower cost?
Ford's Model T was designed with interchangeable parts that fit together with little hand fitting. After the moving assembly line began at Highland Park in 1913, the time to assemble a chassis dropped from about 12.5 hours to about 93 minutes, according to Ford's own records. The savings came from redesigning both the product and the process around each other.
How did compilers shape chip design?
In the late 1970s and early 1980s, researchers at IBM, Berkeley, and Stanford studied what compilers actually produced and found that programs used a small set of simple instructions most of the time. That finding led to reduced instruction set computing, or RISC, in which the chip is built around the instructions a compiler uses. David Patterson and John Hennessy later won the Turing Award for their work on this approach.
More from All-In with Chamath, Jason, Sacks & Friedberg
