Expliseat’s business takes flight with Google Cloud Platform
Expliseat is a French startup that makes the lightest airliner passenger seats on the market. By reducing aircraft weight, their customers use less aviation fuel than before, which leaves the skies cleaner. The rapidly growing technology company needs serious computational power for engineering, designing and testing their product.
The startup makes the lightweight seats from carbon fiber and titanium. They are durable, comfortable, and weigh just four kilograms (half the weight of regular seats), which can save an airline about $450,000 per plane per year through lower fuel consumption. Air Méditerranée became the first customer in July 2014, with several other airlines following soon afterward.
These first sales came at a critical moment in Expliseat’s growth. The company already had 20 engineers working on development alone. They could produce lightweight seats and meet all security and comfort requirements because they had adapted software originally created for the auto industry and relied heavily on computer simulations. Their growing need for IT hardware and services brought Expliseat to a key decision point.
“We were facing the scale issue,” says Benjamin Saada, Expliseat chief executive officer. “The question was, should we buy very big servers and do the calculations in-house, or should we try to work in the cloud?” Saada personally conducted a trial of various platforms to find the best approach for Expliseat. “We decided to move the company to Google Cloud Platform when I saw the performance we needed.”
Google Compute Engine, part of the Google Cloud Platform suite, is particularly useful to Expliseat. It lets companies like Expliseat run large-scale workloads on virtual machines (VMs) hosted on Google’s infrastructure. Compute Engine was a good match for the seat-maker’s massive data crunching.
“All of the calculations for an aircraft seat—how it will react during a crash, all the technical certifications—are made through finite element analysis,” Saada explains. “We go faster with Google Compute Engine. Our engineers can spend more time on the results rather than on the calculations.”
Expliseat had to create with their software vendor a new kind of license and software management system to operate in the cloud. In fact, the application was initially designed to run on powerful in-house servers. “But when we did the first trial on a public cloud, it worked very well,” Saada says. Expliseat’s engineers now run their calculations on Compute Engine’s virtual machines without any special features.
The transition to Compute Engine was quick, and easier than Saada expected. “And compared to the time and the cost to run your own servers, it was faster and cheaper,” he adds. He was also pleased with frequent technical updates—“I expect that Google will be better than I am at improving the systems.”
Expliseat’s computational pace has soared since July 2014, when they began using Compute Engine. “Before, we ran two calculations or tests at the same time, about 10 a week,” Saada says. “With Compute Engine, we can run an infinite number at the same time, with more creativity, more customization and consequently more customer satisfaction. Today, we run 40 or more calculations a week.”
Saada is “very happy” with Google Cloud Platform’s performance. “Since the beginning, we have continually seen new developing tools, new management systems, and new monitoring systems,” he says. “I’m impressed by the pace of improvement, and I’m very excited for the future.”