Nike
Technology is impacting just about every part of our everyday lives – from the way we shop and drive to the way we control the lighting and heating of our houses. Elite sport is no exception. It’s not only the hi-tech world of Formula One either. Even the humble trainer is evolving in line with technological innovation.
When Shelly-Ann Fraser-Pryce headed to the start line of the 100m at the 2016 Rio Olympics, there was something different about her trainers.
Unlike the other shoes on the starting blocks, the Zoom Superfly Elites she was wearing didn’t have spikes screwed in – instead, they were moulded onto one complete plate.
They were also lighter than traditional sprinting trainers and helped her cut her time down in practice.
The trainers were some of the first partly designed by Nike’s algorithms. Now, the technology that helped create Fraser-Pryce’s shoes are starting to trickle down into trainers for the average runner. And that means you could soon be wearing a pair of shoes individually refined by computational design.
The design
“All the texture you see on the shoe is done computationally,” said Bret Schoolmeester, Nike’s Senior Director for global running footwear.
“We started with what we call an envelope, which is just a generic form, and then we add all the texture you see there through generative design.”
The latest running shoes, dubbed the Epic React Flyknit, are the first to use Nike’s new React foam, which is partially made of rubber.
The foam itself is being seen as a competitor to adidas’ Ultraboost and Nike has included more of it on the shoe’s base than in other models.
Where things get really interesting is in the design of the shoe’s sole. The foam on the underside is mostly exposed to the surface below it, but also partly covered by additional rubber protection on the points of highest impact.
Beneath this, the new React foam has a number of grooves, dents, and tracks running along it.
Nike’s machine-led design
These were all designed by Nike’s machine-led design.
“Those tools are able to concept things the human brain can’t conceive and the human hand can’t draw,” Schoolmeester said. The process of computation design involves converting data to structural patterns and telling the system what outcomes it should produce.
“For Rio, it was light and thin, simple. Here’s the plate… go,” Schoolmeester explained. “For this one, it was soft, traction, and stability.”
Each different size of Epic React has had its pattern graded. There are subtle differences in the depth of groves and placement in texture between a male size ten shoe and a female size five, Schoolmeester said. Changing the geometry of the foam allegedly makes it softer on impact with the ground.
Computational design is becoming popular across multiple industries. The technique often reduces the weight of materials, creates new complex structures and is speedier than using human designers.
The field itself isn’t new – Carnegie Mellon University has been researching the areas since the 1960s.
However, what is making it popular now is reliable machine learning systems, greater quantities of data, increased processing power, and faster manufacturing.
In his annual tech and design report, John Maeda, a former MIT professor and head of computational design at Automattic, argued computational design means products can continually evolve and the system works for products that are used by ‘hundreds of millions of people’.
Nike has already used the method in a T-shirt and is recruiting for a product director to further develop ideas in the area. The technique also has the potential to change the role of a designer.
“Our designers and now software programs don’t need a break,” Schoolmeester said. “They can have a different idea every minute.”
