Nieke Roos
16 November 2021

High-tech systems are capital intensive, so keeping them operational is of paramount importance. Neways has developed a remote performance diagnostics methodology that can significantly speed up the build time and reduce unplanned downtime.

To remain at the leading edge, the high-tech industry is knuckling down on system performance, time to market and total cost of ownership (TCO). Each new machine generation, however, not only brings a performance increase but also a proportional rise in complexity along with it. This, in turn, raises the likelihood of integration issues arising when putting internally and externally developed subsystems together, as well as the chance of stoppages or faults occurring during operational use. Either way, productivity goes down and TCO goes up.

With its remote performance diagnostics methodology, Neways is assisting high-tech companies to help keep integration issues and unplanned downtime in check. “The approach improves the interaction between subsystems, minimizes the margin of error and allows for remote health status monitoring,” explains creator Ron Schram, system architect at the Son-based provider of electronic manufacturing services. “By designing electronics more smartly and analyzing potential problems in advance, system quality and availability can be increased and serviceability can be more targeted.”

Neways RADS
The methodology is based on four pillars: reliability, availability, diagnostics and serviceability. Credit: Neways

According to Schram, the methodology brings great benefits, both in the time needed to build and validate a system as well as in the mean time to repair. “Having diagnostic information available in relation to the entire system reduces the time customers have to spend troubleshooting integration issues and speeds up the build time. The use of remote performance diagnostics has also been shown to reduce unplanned downtime from hours to mere minutes.”

RADS

“At Neways, we design boards, racks and cabinets, assemble them, test them and deliver them to our customers, who then integrate them in their machines. When everything works, the system is disassembled and shipped in pieces to the customer’s customer, where it’s put back together again,” Schram says, describing the development process in a nutshell. “Along the way, all kinds of things can happen with our units causing them to malfunction at some point, in which case we get called to figure out the problem and fix it.”

In Neways’ experience, the vast majority of errors arise during integration at the customer and re-assembly at the customer’s customer. “This got us thinking: what if we could detect them right away? We then came up with this methodology. By adding remote performance diagnostics to our units and the supporting infrastructure on their end, our customers can achieve huge cost savings with a minor investment.”

The methodology is based on four pillars. “Internally, we refer to it as RADS: reliability, availability, diagnostics and serviceability,” Schram sums up. “It starts with our units being reliable, meaning that they have a high MTBI, mean time between interrupts. Using extensive analysis and testing, you need to design them in such a way that you minimize the chances of errors.”

When a component does happen to break down, the system needs to remain functional for as long as possible. “One way to achieve this availability is through redundancy. By using multiple power supplies and redistributing the load when one of them stops working, for example. Or by passing on a set of fallback values instead of the output of a faulty sensor. A system can also be unavailable simply because a board has been inserted askew or even upside down, so you need to make sure that that’s impossible.”

Both the reliability and the availability can be monitored using diagnostics. “You can do all kinds of health checks. Does a board contain the latest firmware version? Is it properly connected to the rest of the system? Is it communicating with the optimal data rate? With a motor or a fan, you can look at the power usage to detect imminent failure. A lot of the hardware checks can be performed by an onboard FPGA.”

Of course, any problems need to be fixed as soon as possible, with as little effort as possible. “Components should be easily accessible so that when they break down or are about to, they can be replaced quickly – in seconds rather than hours. Preferably, they’re hot-swappable, meaning that they can be substituted without having to power down the whole system. Serviceability also includes the ability to provide the specific location of a faulty component. Being able to pinpoint a service engineer exactly to where he needs to be saves time and thus money.”

Neways Ron Schram
“The investment is peanuts compared to the total bill of materials while the return is massive,” states Ron Schram. Credit: Neways

Reticence

The methodology requires bilateral involvement, Schram points out, from Neways as well as the customer. “We can equip our units with all kinds of self-tests, but we’re nowhere without a supporting infrastructure. To actually achieve reliability, availability, diagnostics and serviceability, customers need to do their part and create the environment to capture and process the data we provide – using artificial intelligence, for example. Once the infrastructure is in place, we can start introducing backward compatibility and gradually replace older units that don’t provide remote performance diagnostics with new ones that do. It’s a step-by-step plan.”

Although they all feel the need for more predictive maintenance, it takes some convincing to get high-tech companies on board. “We see some reticence,” Schram notices. “In the end, it works, so why spend money to make the infrastructure suitable for diagnostics? We take them on a journey to the big picture and try to explain that the investment is peanuts compared to the total bill of materials while the return is massive. And sooner or later, components and even boards need to be renewed anyway. Why not spend 1-2 percent extra and include remote performance diagnostics?”

The launching customer was a bit hesitant as well at first. “When we pitched the methodology to them five years ago, they didn’t immediately embrace it, but in the end, the system reliability, the low downtime, the TCO reduction and a satisfied end customer convinced them. Now, it’s been implemented in their latest product and we’re working on the next generation together.”

This article was written in close collaboration with Neways.