Auto Industry Wrestles with Need to Build ‘Intelligent’ Factories

August 25, 2016 by Jerry Hirsch, @Jerryhirsch

Inside a sprawling Nissan auto factory in rural Mississippi, a robotic cart transports a massive rear axle to intersect a partially built Titan XD pickup truck inching forward on the assembly line.

With mechanical patience, the cart pauses as the truck frame moves into place. It slowly rolls to a factory worker, who uses a semi-automated hoist to bolt the axle onto the truck frame. The process represents a sophisticated melding of human skill, robotic technology and complex inventory-management software.

Yet, like every other auto factory in the world, this Nissan plant is struggling to keep up with technological advances. The problem, explains Steve Marsh, Nissan’s vice president of manufacturing at the Canton, Miss., assembly plant, is that the cars are starting to be smarter than the factory.

“If we are not careful, the vehicles we build will have more technology than the plants where they were built,” Marsh said. “Factories need to keep pace.”

It’s a surprising statement from a veteran plant manager who oversees one of the most advanced automotive factories in the U.S. The 4.7-million-square-foot plant employs 6,400 humans and 1,400 robots to seamlessly build eight different vehicles, including four versions of the Titan and Frontier pickup trucks, two models of the NV van, the Murano sport utility vehicle and the Altima sedan.

But even so, his plant is falling behind the technological curve.

“Robot technology has moved forward over 20 years, and we have not caught up,” Marsh said.

The typical auto factory like Canton, for example, maintains a safety fence around many types of robots. Technological advances will make that unnecessary.

“Now there are robots that people can work next to or in tandem with without a fence or other safety barrier,” Marsh said.

Protective fencing gobbles up valuable plant real estate that could be used to produce more vehicles and adds to production costs.

“Fencing and safety equipment take up space and cost money,” Marsh said.  “They can double the cost of a $70,000 robot.”

But the challenges ahead are more than physical.

Robots and Humans Build the Titan XD Pickup Truck:

Factories need more-sophisticated equipment that will be able to test complex electronic systems on vehicles packed with electronics and that are rapidly transforming to digital devices from machines.

“The equipment at factories to test vehicles needs to be significantly more advanced than it is today,” Marsh said. “If we have vehicles that can steer and stop themselves, we can’t have any variability in how they operate. The vehicles have to be foolproof.”

How to retool factories to improve productivity and manufacture intelligent vehicles is an industrywide issue, Marsh said.

Other auto industry executives agree.

“The true problem, the true difficulty, and where the greatest potential is, is building the machine that makes the machine­ — in other words building the factory,” Elon Musk, chief executive of Tesla Motors, told shareholders at the Palo Alto electric car company’s annual meeting this year.

One problem, he said, is that cars are assembled at about the pace of a “tortoise” walking.  There’s potential to dramatically increase production speed with automation.

Speaking to analysts this year, Musk said he envisions a factory that “might look like a giant chip pick-and-place machine or a super high-speed bottling or canning plant, and you really can’t have people in the production line itself.”

Once humans are involved, he said, production drops to “people speed.”

“There’s still a lot of people at the factory,” Musk said, “but what they’re doing is maintaining the machines, upgrading them, dealing with anomalies.”

“We are going to design a factory like you would design an advanced computer,” Musk said.

While the auto industry is starting to talk about a dramatic rethinking of how it manufacturers cars and trucks, it’s not taking much action yet, said Arun Kumar, a director in the automotive practice at AlixPartners, a global consulting firm.

“There is a lot of talk about how workers and machines can work together, that’s sexy,” Kumar said. “But you still have to get down to the brass tacks.”

For the most part, industry management is still stuck in a “1980s and ’90s mind-set,” he said.

“There is a talent shortage,” Kumar said. “You need more people who are digitally capable, who can use big data and analytics, connecting to the cloud and developing the human machine interface.”

Nissan Titan XD Pickup Truck

Nissan Titan XD pickup truck rolls off assembly line at automakers Canton, Miss., factory. (Photo: Nissan)

The industry isn’t investing enough resources in these areas, he said.

AlixPartners estimates U.S. manufacturing industries collectively need to invest $60 billion to $80 billion into the so-called internet of things by 2020. Kumar doesn’t have an estimate for the auto industry’s share of that total but said it will run into the billions of dollars.

Automakers have to do a better job integrating existing digital technology into their manufacturing systems, Kumar said.

“It is not like you have to wait for someone to invent it,” he said. “With the internet of things, you can put cheap sensors at different points in the network to tell you what is really going on.”

Easily collectible data, for example, can help workers analyze why a specific machine is more likely to fail during the second shift of a production day. Once a factory figures that out, it can retrain workers or change the manufacturing process to improve reliability, he said.

“We call that the predictive maintenance approach,” Kumar said. “You can predict when a machine might fail based on the temperature or the viscosity of the oil in it.”

There are also new ways factories can reduce costly downtime on their assembly lines.

“If a worker finds a machine is slowing down or failing, there are tools in the marketplace today that allow a worker to pick up a tablet, talk to someone thousands of miles away and do a diagnostic review,” he said.

The auto industry has to move faster to adopt these types of production tools.

“The industry improves productivity at about a 4 percent annual rate, but it will slow down unless it adopts a more digital approach,” Kumar said.

Eventually, the industry can retool with the science-fiction type technology that Musk and others are contemplating, but that’s still well into the future.

“In the short term you will see the status quo,” Kumar said. “The robotics tech is not mature enough to drive a step change, but it will happen with mini-robots that talk to each other and work together with humans and intelligent conveyor belts.”


Subscribe to our mailing lists

Choose one or more topics: