Kathy Gibson is in Parma – High-performance computing makes it possible for racing car manufacturer Dallara to quickly try new designs and make mistakes in the laboratory before they invest in actual production.
This allows designers to be truly innovative, says Dr Andrea Pontremoli, CEO and GM of Dallara Group.
If designers don’t make mistakes, he says, they are simply doing what they already know.
“Eveything we do is based on one key item: the only way to be innovative is to have the ability to make mistakes. If you don’t make mistakes you cannot be innovative.
“So if you want to do something new you have to accept mistakes. The way to do this without bankruptcy is through simulation.”
Dallara uses its Lenovo supercomputers to perform realtime simulations that quickly show up mistakes in the process, and give the company the opportunity to fix these before going into production.
Dr Pontremoli came to Dallara from IBM, where he was first a customer engineer and later CEO of the southern Europe region. In 2007 he joined the family-owned Dallara as CEO and partner.
The company is not particularly well-known in the market, operating as a business-to-business supplier to some of the biggest racing car companies in the world – many of them in the 100-km area of Italy colloquially know as Motor Valley.
The business is complex, he says, because of the speed at which the industry is changing. “The name of the game is speed: both the economy and the way of working are all changing.”
Dallara’s business focuses on three areas: lightweight construction using carbon fibre; aerodynamics; and vehicle dynamics.
The company is able to take a new car from concept to production in just nine months, with the ninth month used for production.
The only way to do this is to use supercomputer capabilities to simulate every part of the process, Dr Pontremoli explains.
“We use supercomuters to siimulate advanced compsite materials, aerodynamics and vehicle dynamics.”
He points out that just 15% of a Formula 1 car’s performance is driven by the engine. A more significant 35% is derived from the weight; and 50% from the aerodynamics.
“We don’t work on the engine, but are responsible for the other 85% of the car’s performance.”
Simulation is used to find the perfect material composite that will keep the weight down. An 850kg car is a lot easier to move and stop than a 1 600kg car, which is the typical weight of a small commercial sedan, says Dr Pontremoli.
Aerodynamics is the science of adding weight to the vehicle without adding mass that has to be moved, or can affect performance in turns.
“We add 820kg of weight in air,” Dr Pontremoli explains. “We use simulation to understand the effects of aerodynamics before we build the car.
“Supercomputing allows us to reduce the time and cost of producing the most aerodynamically-efficient cars.”
Vehicle dynamics is the sum of all the other processes and studies the end result.
“Using simulation, the driver can drive a car that was never built.”
Dr Pontremoli cites the example of designing a new hybrid car: there are many options just for engine placement and interaction.
“Where do you put the engines? How many should you have? How should they work together? If you have to build all possible architectures, it will cost a fortune and take a lot of time and effort.
“The simulator lets you test all of the alternatives and select the one to develop.”
Simulation does present some technological issues that Dallara has had to overcome. “This biggest of these is related to realtime, and the ability to give realtime responses to the driver,” says Dr Pontremoli. “When he turns the steering wheel, the system can’t take two seconds to respond. For the driver, it needs to be seamless, as it he was on the racetrack.”
Dallara has been able to solve this and other issues, and has developed the mathematical models needed to simulate all the car parts.
Importantly, they are developed in open source and so they can work together with models deloped by other players in the industry. “Because we have a common language, we can keep them together and run them together,” Dr Pontremoli says.
Simulation is such a powerful tool that Dallara has even skipped prototyping and gone straight to production on the strength of simulation results.
“You can see the difference in a company that works the way we do to a company that uses only production to test,” Dr Pontremoli says.
It is almost impossible to quantify how much time and money Dallara saves by using simulation, because the entire process – with the design innovations that the industry demands – would be almost impossible to do in the time that’s required.
“The power of the technology allows us to make more mistakes and try more than we could before, in a given time.”
Artificial intelligence (AI) is used to optimise the level of data retrieved from vehicles in simulation and in action on the track, and to understand some of that data.
“We have all the data from our cars racing around the world,” Dr Pontremoli points out.
“We also use AI to recognise given patterns, to extract the patterns and avoid simulating something we have done already.”
Data analysis and innovation has become so key to Dallara’s operations that is has helped to massively grow the company over the last 13 years, but also to diversify.
Back in 2007, operations were 100% focused on racing car design; today, that accounts for just 60% of the business, with consulting to other companies in the global motor racing industry accounting for a massive 40%.
While it’s changed its own company to reflect industry dynamics, Dallara has also changed the way customers typically deal with their IT suppliers.
“We grew up with the idea that companies need to understand technology and how to apply to our business model. But we have no time to do this anymore.
“I have to assume that the technolgoy to do what I want to do in my business exists, and I need to be able to tell my supplier what I want to achieve – and they must make it happen.”
Per Overgaard, executive director of the Lenovo Data Centre Business Group EMEA, confirms that this is the new way of doing business for many companies.
“Customers now have a particular issue they want to solve and they come to us for that solution. As a result, the world is flipped and we are now integrating customer questions into our R&D.”
In a world where business needs are changing all the time, the relationship between IT suppliers and their customers has to change.
Alessandro de Bartolo, GM of the Data Centre Group at Lenovo Italia, points out that a working relationship like the one Lenovo and Dallara have helps the company to grow.
“The requirements we have from them are always very challenging,” he says. “And, because they are a leading company in the world of motorsport, support is always a challenge.
“But, because of this, we have had to grow our own capabilities.”