Smart Factory Developments Create New Demands on Power

When a product defect is discovered in a traditional assembly line, an inspector hits a red emergency stop button. Engineers and managers rush in to find the problem in the line and then troubleshoot a solution. This brings the entire operation to a halt for an indeterminate amount of time. Each delay has the potential to derail cost estimates and lengthen production schedules.

A key goal of Industry 4.0 is the realization of self-monitoring assembly lines, where inspection and correction occur automatically, without costly production delays.  

In this scenario, vision systems can identify quality issues without direct human involvement. Machine learning (ML) modules analyze the visual data in real time, calculate machine adjustments and correct the course for the next product on the line.

This industrial ‘auto-correct’ capability may have many guises in the future. The same sophisticated ML that infers facial recognition can be trained to spot flaws in the layers of 3D printed parts. Laser scanning and measurement can detect parts that are out of alignment. Self-correction could evolve incrementally as additional hardware attachments to existing robots or cobots, autonomous factory vehicles (AFVs) or Industrial Internet of Things (IIoT) devices are introduced.  

The prospect of no longer having to press pause on the conveyor is an attractive business objective. In the fully automated factory, production time becomes much more predictable, with less risk of cost and schedule overruns. The risk for human error is diminished, too. AI may soon become capable of optimizing quality based on its own rules, far surpassing the limits of manual inspection. 

All efforts toward enabling a truly smart factory have a common denominator: the need for new power infrastructure. 

Smart factories introduce a higher demand for electrical power for both the larger machinery performing production tasks and the new extensive information network.

This layer of communication and control includes wiring and devices, like cameras, sensors, actuators and control units. A self-correcting assembly line might also require a range of electrical upgrades, such as transformers, switchgear power supplies and power distribution panels, to accommodate the additional load. 

Ideal smart factory signal-and-power components would combine a ruggedness to survive on the floor with the speed and reliability attributes of data center infrastructure.

Industry standard organizations like Underwriters Laboratory (UL) and National Fire Prevention Association (NFPA) are already anticipating these requirements and other emergent issues. For example, more powered devices on the floor introduce more frequencies that can interfere with machine performance. This means sensitive electronics such as those found in servo motors are particularly susceptible to disruption from nearby devices in the environment. 

Data centers often take extra measures to ensure power quality. Forward-looking industrial facilities are following suit, generating current that is smoother and more reliable. 

A smart factory with a greater emphasis on power quality might feature new types of components. Capacitive banks, for example, are employed to eliminate fluctuations in supply lines. On the demand side, variable frequency drives give robotic motors a softer start. Rather than an intense drag on power as they come up to speed, these motor drives consume power evenly over a few seconds.

If an assembly line without interruption is the goal, then it is imperative that the communication, control lines and data processing keep running. As is frequently the case with data centers, critical computing units may each have their own dedicated battery backup. In addition to batteries, both factories and data centers are also protecting against the effects of dropped power by incorporating on-site sources of energy.  

As AI-driven self-correction gains hold, factory power managers will have more decisions to make. When to switch power supply will merely be one of them. Fortunately, AI-driven data analysis provides factories with something else: an enhanced ability to monitor and manage power. This will be welcomed by the more than 800 design engineers who participated in a Molex power survey. When asked about their main priorities for designing or implementing a power system, they responded as follows: increase energy efficiency (74%), reduce costs (64%) and increase monitoring of power system status (53%). 

Real-time AI-driven diagnostics works for more than correcting product flaws. A smart factory of the future would benefit from control systems that can track voltage and current levels throughout the operation.

By monitoring each production activity consuming power, managers can prevent total usage from exceeding physical or compliance limits and predict the power requirements of activities based on previous data. 

Power diagnostic tools would enable managers to maximize uptime and schedule preventative maintenance during predictable transitions. A higher-resolution view of power consumption shows operations managers how to add more production activities given the limits of available power. 

Real-time power monitoring also enables better load balancing, reduces transmission losses and improves the resilience and flexibility of the power distribution system.  

In the recent Molex survey of power system engineering professionals, the number one challenge for industrial applications was power management by a large margin — cited by more than a third of engineers and managers working in the industrial space. Power management was similarly the top challenge for professionals in data centers, by almost the same margin: 40-percent. 

The convergence of automation, data analytics and advanced technologies in the factory of the future presents a unique set of challenges for power management as well as opportunities for new levels of autonomy and control. 

Manufacturers must navigate the risks associated with EMI, especially with the increasing reliance on smart tools. Embracing self-adjusting assembly lines powered by vision systems offers higher precision, quality and productivity. And real-time AI-driven diagnostics for power infrastructure empower factories to optimize their operations and ensure compliance while minimizing downtime. 

As the factory of the future continues to take shape, harnessing the power of data becomes essential for unlocking new levels of efficiency, productivity and innovation in manufacturing processes.