Maximo monitors the health of production machinery and tooling by using data to predict when critical maintenance is needed.
The upcoming extended-range Ram 1500 Ramcharger pickup won't look much different from its gasoline-powered counterpart. But under its fenders, the vehicle will use a flexible chassis with many different parts and, of course, a completely different drivetrain, wiring harnesses, cooling packages and more.
Now add in the myriad models, trim levels and options and the fact that both the Ramcharger and the internal combustion engine-powered 1500 pickups will be built on the same production line, and you have immense manufacturing complexity. Over at rival Ford, there are more than 400,000 permutations of F-Series pickups.
This is the type of complexity where artificial intelligence such as IBM's Maximo software can help automakers reduce production line-stopping mistakes that cost money and ding quality.
IBM's software experts are working to expand Maximo's capabilities to help ensure the right parts are installed on vehicles and to deliver early warnings when production machinery is in need of maintenance.
"AI-based algorithms can digest masses of data from vibration sensors and other sources, detect anomalies, separate errors from background noise, diagnose the problem, and predict if a breakdown is likely or imminent," a McKinsey & Co. report on automotive uses for AI found.
That could result in a "more than 20 percent increase in equipment availability," according to the consulting firm, as well as up to 25 percent lower inspection costs and up to 10 percent lower total annual maintenance costs.
For years, this suite of vision-based AI tools has helped automakers and suppliers reduce manufacturing defects on assembly lines right where they occur. If a part is missing or incorrectly oriented in a transmission, for example, the line comes to a stop and the problem can be quickly fixed on the spot. Stamping out defects where they occur saves both automakers and suppliers money.
Some version of Maximo, which debuted in the mid-1980s, is in use at more than six of the world's 10 largest automakers, according to IBM.
Now, IBM has been expanding the software's capabilities to new frontiers to address growing manufacturing complexity and to help plant managers avoid unplanned-for downtime caused by faulty machinery. Maximo monitors the health of production machinery and tooling by using data to predict when critical maintenance is needed.
Brett Hillhouse, Maximo global industry leader for automotive and electronics, said most auto companies are using Maximo for enterprise asset management. The software helps automakers and suppliers increase plant uptime, lower costs, manage quality and extend the life of equipment.
Enterprise asset management is very nuanced. And it involves more than just measuring how quickly tooling is wearing out or if a machine's calibration is slipping.
"Traditionally, you would do inspections. But now with AI, what we've been doing is applying it more in asset performance management," said Michael DeSabaris, Maximo reliabilities strategy product manager. "We can actually monitor the condition in near real time of the health of that asset. That allows you to intervene early so you can actually predict when a failure is going to happen."
That reduces unscheduled downtime, he added.
Hillhouse said one of Maximo's capabilities is constantly comparing parts that are in specification in an effort to prevent a defect before it happens. One example involves a metal part stamped by a die in a huge press, such as a fender.
"Even though it might still be in specification, if it is starting to trend towards the end of a specific tolerance that is allowed, that should be a trending item that we know about," Hillhouse said. "We can do correlations between parts. Where am I on the calibration of that unit itself in terms of wear and tear of that specific die?"
Another benefit of using Maximo to monitor the status of production equipment is that it allows plant managers to plan downtime for maintenance so that it doesn't disrupt production schedules, DeSabaris said.
"As the function of an asset starts to degrade ... you are intervening early, taking action right away at your convenience," he said. "It's very efficient and it saves you money because you can plan so you don't have that downtime."
Many machines used in manufacturing generate pressure, temperature and vibrations that create a specific signature that can be monitored. "If something starts to change, that would prompt an investigation that would show that, maybe, the mold is starting to wear, or something within the machine is wearing," DeSabaris said.
A big push at IBM is to tie Maximo analytic capabilities in with production machinery software made by other companies.
"The manufacturing execution systems guys are always focused on defects," said Hillhouse. "They are now realizing defects often have some correlation back to the equipment, so we are working with some of the manufacturing companies, like Siemens, on how we tie our systems together to work better."