learning by trial and error can be exciting. the lessons learned deeply ingrained. learning by failure is a remarkably good way of gaining knowledge. it is part pf learning.

With a quite respectable product to present, I set off around the world to start selling it properly. It was time spent away from designing, but it was to teach me, above all else, that only by trying to sell the thing you have made yourself, by dealing with consumers' problems and the product's failings as they arise, can you really come to understand what you have done, to bond with your invention and to improve it. Conversely, of course, only the man who has brought the thing into the world can presume to foist it on others, and demand a heavy price, with all his heart.

My big mistake had been presenting the same craft to each customer and telling them: 'This can be adapted to suit your needs' If someone wanted a diving boat I would explain that it could be fitted with compressors, heaters and a very slow diesel engine. If an oil company wanted a crew bus, I would tell them that suitable seating and a faster engine could be fitted. To the military I said I would bullet-proof the sides and engine. To constructors in search of a bridging tug I said, 'Special buffers? High power engine? No problem.' I convinced not a single one of them.

The root principle was to do things your way. It didn't matter how other people did it. It didn't matter if it could be done better. The Ballbarrow was not the only way to make a wheelbarrow that didn't get stuck in mud - but it was a way. The trick is not to keep looking over your shoulder at others, or to worry, even as you begin a project, that it is not going to be the best possible example of its kind. As long as it works, and it is exciting, people will follow you.

Like Brunel, Jeremy Fry did not, when an idea came to him, sit down and process it through pages of calculations; he didn't argue it through with anyone; he just went out and built it. as far as he was concerned, with enthusiasm and intelligence anything was possible. It was mind-blowing. No research, no 'workings', no preliminary sketches. If it didn't work one way he would just try it another way, until it did.