`Teach engineers to exploit statistics’

Statistics should be integrated into engineering undergraduate courses as a tool for improving reliability and reducing variability in the manufacturing and design process, said a senior Ford engineer this week. Delivering the Royal Academy of Engineering 1999 Engineering Manufacturing Lecture, Richard Parry-Jones, Ford group vice-president for product development and quality, said engineers stuck to a […]

Statistics should be integrated into engineering undergraduate courses as a tool for improving reliability and reducing variability in the manufacturing and design process, said a senior Ford engineer this week.

Delivering the Royal Academy of Engineering 1999 Engineering Manufacturing Lecture, Richard Parry-Jones, Ford group vice-president for product development and quality, said engineers stuck to a deterministic, physics-based approach to engineering design.

But statistical methods could be more powerful for improving processes where the interaction of many factors were not well understood.

He appealed to `the most influential figures in our profession here tonight’ to ask for `statistical engineering methods to be taught and embedded in our institutions’ undergraduate curricula and requirements for professional experience.’

He added that statistics should be taught by engineers in the context of design and engineering, not as a separate course.

Parry-Jones said statistically-based methods of experimenting were counter-intuitive to many of Ford’s professional engineer and engineering graduate recruits: `We have to teach them,’ he said.

Parry-Jones said Ford had effectively used statistical techniques on the production line and in design to reduce the variability inherent in manufacturing.