United StatesConsumer goods2019 – ongoingCallaway (Topgolf Callaway Brands)
Machine learning to design driver and iron faces
Callaway said machine learning cut some club design tasks from four to five weeks to a couple of days
Problem
Club development at Callaway typically took 12 to 16 months and produced only three or four prototypes. Engineers set face designs by hand, and testing each idea was slow.
Approach
Callaway fed design parameters and 20 years of golf shot data into machine learning run on a supercomputer of about 1,000 cores. The software worked out the face design on its own. For the Flash driver it produced a thickness pattern with ripples and ridges that engineers could not fully explain but that gave more ball speed.
Results
Takeaway
Machine learning let engineers test far more designs in the same time. The speed claim came from Callaway's R&D chief in a 2020 interview.
Categories
- Size
- Large enterprise
- Industry
- Consumer goods
- Business function
- IT and R&D
- Innovation type
- Product and service innovation
- AI technology
- Predictive analytics and MLOptimization
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
Popular Science described how Callaway used machine learning and a roughly 1,000-core supercomputer to design the face of its Flash driver. Alan Hocknell, senior vice president of R&D, said tasks that took four to five weeks now take a couple of days. The article gives typical development times of 12 to 16 months.
Full details are in the original sources.
