Thursday, February 25, 2010

The two experimentation loops in an innovation: the story of Watt-Boulton steam engine

When we think of an innovation like say, the steam engine, we say, “James Watt invented the steam engine, manufactured it and made money”. This linear thinking from “idea” to “implementation” to “cash” (or “crash”) is too simplistic and hides an important non-linear process in the story: experimentation. For example, it doesn’t tell us that it took ten years of experimentation to demonstrate steam engine to public. Question is: what kinds of loops are involved in an innovation story? Let’s see two experimentation loops typically involved in the lifecycle of any idea using the Watt-Boulton steam engine story. We will first see the story in brief.

1. James Watt was a mathematical instrument maker to Glasgow University. In the winter of 1763-64 John Anderson, a professor of natural philosophy brought a small model of Newcomen engine (the prevalent engine model) for repair. Watt was perplexed by several aspects of the engine including the unexpectedly large quantity of steam consumed.

2. Watt started experimenting with the engine during the free time he got from his regular duties. After a year of experimentation he recognized that there is a paradoxical deficiency in the Newcomen engine. To utilize the steam efficiently, during the unpowered-stroke, the cylinder had to be kept at 100O C. But during the power stroke considerable cooling water had to be injected into the cylinder, this tended to cool the cylinder below 100O C.

3. Then one day, when Watt was strolling on the Green of Glasgow early in 1765, came his famous insight: he would condense the engine’s steam not in the operating cylinder, as Newcomen and all his followers had done, but rather in a separate condensing vessel to which it would be drawn by a pump or by other means.

4. For the next 10 years, Watt was to perform several experiments on cylinders of various sizes starting with small ones such as 1.4 inch, 1.75 inch, 6-7-8 inch and later on larger ones such as 18-38-50 inch etc. He also experimented with various materials such as brass, sheet iron, copper and wood. He also filed a patent in 1768. His work was funded by three investors, first by his friend Prof. Joseph Black, then by a businessman Dr. John Roebuck and finally by another businessman Matthew Boulton. Boulton also successfully re-negotiated the patent’s timeframe and got it extended by 25 years (till 1800).

5. In 1776 the first successful prototype was demonstrated with a 50-inch water-pumping model for a mine and a 38-inch model for blowing John Wilkinson’s blast furnaces. Both engines ran well and lead to favorable publicity.

6. From 1776 the firm began to install engines elsewhere. The firm rarely produced the engine itself: it had the purchaser buy parts from a number of suppliers and then assembled the engine on-site under the supervision of an engineer from their firm. The company made its profit by comparing the amount of coal used by the machine with that used by an earlier, less efficient Newcomen engine, and required payments of one-third of the savings annually for the next 25 years. In 1779 the firm hired engineer William Murdoch, who was able to take over the management of most of the on-site installation problems, allowing Watt and Boulton to remain in Birmingham.

7. Between 1775 and 1800 the firm produced approximately 450 engines. It did not let other manufacturers produce engines with separate condensers, and approximately 1,000 Newcomen engines, less efficient but cheaper and not subject to the restrictions of Watt's patent, were produced in Britain during that time. Watt, who started in business by borrowing, left an estate of £60,000 at his death in 1819. Boulton’s estate in 1809 was £150,000.

This figure shows us the two experimentation loops: (1) idea-to-demo loop and (2) demo-to-cash loop. Note that the first loop is a decade long and it involved two investors (Black and Roebuck) running out of money. After the demo, it took another three years to stabilize the business model.

Innovation is more like a marathon than a sprint with a few circular tracks in between!

Source: Invention and Innovation in the Watt-Boulton Steam-Engine Venture, F. M. Scherer, Technology and Culture, Vol 6, No 2 (Spring 1965), pp 165-187. (not free)

Also see:

Learning about the innovation process through 20 point stories

Assessing capacity of innovation engine using CEO model

Experimentation matters by Stefan Thomke (My favorite book on systematic experimentation)

1 comment:

  1. Another excellent anecdote with a pithy message. Keep it up!