1 / 25

Editing as an Evolutionary Algorithm: Selection and Exploration

Discover how John Deere engineers used evolutionary algorithms to solve production challenges in a complex manufacturing environment, mirroring nature's evolutionary processes. Learn about the power of selection and exploration in guiding the creation of efficient manufacturing schedules. Explore the iterative nature of evolution and editing, where mistakes lead to new ideas and improved outcomes. Uncover the parallels between evolutionary algorithms and the creative writing process, showcasing the potential of embracing iterative improvements and embracing the unknown.

tron
Download Presentation

Editing as an Evolutionary Algorithm: Selection and Exploration

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Editing as an Evolutionary Algorithm: Selection and Exploration

  2. The new John Deere 7290R.. an amazing advance in tractor technology...

  3. A modern manufacturing dilemma John Deere makes tractors, planters, and lots more The manufacturing complexity is now overwhelming ... thousands of products... billions of possible combinations...

  4. Problems In the late 1980s, in trying to fill hundreds of custom orders each day, John Deere engineers faced: Horrific traffic jams in the plant Machines in demand at more than one place at a time Others remaining idle with nothing to do

  5. What to do? Engineers couldn't find efficient production schedules, what to build, in which order. But eventually, Bill Fulkerson had a creative idea: what's more intelligent than the human brain?

  6. Answer: Evolution

  7. Evolution: “descent with modification” - a messy process, depends on selection and exploration

  8. Can evolution be used to design... manufacturing schedules?

  9. Any sequence for running the plantis akin to a “DNA” strand Make tractor Make planter Make ?? Maintain machines Schedule 1 Schedule 2 Schedule 3 Time

  10. Start with a “population” of random schedules Make tractor Make planter Make ?? Maintain machines Schedule 1 Schedule 2 Schedule 3 Time of day None of these will be any good, but let them evolve...

  11. Use computation to let the populationEVOLVE one generation to the next

  12. Calculate the “fitness” of each schedule..faster means fitter Schedule 1 100 hours Schedule 2 200 hours Schedule 3 360 hours Let SELECTION act: fitter schedules have more offspring in the next generation

  13. Also include EXPLORATION Allow some point mutations: Allow “sexual recombination” through crossover: Cut here & swap

  14. Results? Early on, all schedules perform poorly After evolution overnight, some schedules outperformed anything the engineers could design They work -- but NO ONE really understands why!! Early on, all schedules perform poorly After evolution overnight, some schedules outperformed anything the engineers could design They work -- but NO ONE really understands why!!

  15. selection + exploration + iteration = A powerful algorithm for finding solutions to extremely complex problems

  16. Like evolution, writing is also a messy, iterative process

  17. My book is smarter and more creative than I am Why???

  18. Pamela Druckerman “A large part of the creative process is tolerating the gap between the glorious image you had in your mind, and the sad thing you’ve just made.”

  19. Next comes... revision... editing...

  20. Evolution: “descent with modification” - a messy process, depends on selection and exploration

  21. The evolutionary dynamics of editing: One draft gives rise to the next through: Selection: - you keep the better, reject the worse Exploration: - you expand on incomplete thoughts - make wrong thoughts less wrong - try out new ideas - silly, half-crazy, stupid ideas all good, as raw matter for further selection

  22. Old ideas launch new ones

  23. Stephen King “I write to find out what I think.”

  24. & evolution My book is smarter than I am Why: co-authored by evolution!

More Related