A primary manufacturing goal is to maintain high product quality. For many enterprises, however, this objective still seems hardly achievable. Too often, product quality issues are uncovered only when a product fails in testing, or worse, when a customer makes a return or triggers a recall.

A common cause behind reduced product quality is faulty equipment, which has not been properly maintained or calibrated. Manufacturers are increasingly turning to IoT-driven machine condition monitoring, which helps reveal equipment issues that can affect the quality of products so they can be fixed before things get worse.

The IoT-driven approach to product quality control

Condition monitoring enables product quality control by detecting combinations of equipment health, such as spindle vibration frequency, engine temperature, cutting speed, and ambient parameters, such as temperature and humidity. Combined, these parameters can cause deterioration in the quality of a product output.

A historical data set that contains equipment condition records gathered through a time period (say, a year) is combined with the data bout product quality deviations and context data (for example, equipment maintenance history) from either ERP, PIMS, or DCS systems. The combined data set is then fed into advanced machine learning algorithms, which can then detect causal correlations in the incoming data records. Uncovered correlations are reflected in predictive models, which are then used to identify combinations of equipment condition and environmental parameters that can lead to product quality issues.

For example, in pulp processing some of the quality issues include deviations in the concentration of dissolved alkali. The machine learning component of IoT detects hidden patterns in the data and states that a higher concentration of alkali stems from a deviation in two process parameters: reduced processing temperature and increased white liquor flow.

Use cases across industries

Manufacturers across industries can leverage IoT for monitoring the condition of machines and controlling the quality of products and components manufactured on them. Here are a few examples:

Pulp and paper

In the pulp and paper industry, IoT allows monitoring the condition of rollers in paper machines. A defect of just one roller bearing can significantly affect the quality of the produced paper and cause fluffing and changes in paper thickness. Monitoring the condition of roller bearings with vibration sensors is enough to avoid a large percentage of quality issues. Vibration sensors on each end of the roller continuously gather real-time data about the roller health and relay it to the cloud software. If a roller does not function properly, an IoT solution alerts an operator.

Electronics

In electronics, in the process of mounting semiconductors on circuit boards, tiny puffs of air are used to direct the placement of chips. The placement machines are calibrated according to current environmental conditions (temperature, humidity, etc.). A minor change in, say, temperature parameters generates a heat profile that can cause placement defects. Temperature and humidity sensors are used to monitor the environment, in which machines operate. Once a change is detected, a machine can receive a command to make calibration adjustments to meet the quality standards.

Steelmaking

In the steel industry, IoT helps detect equipment issues that affect the quality of steel during the metal forming process. During the process, a slab – an output of casting – is reheated and run through the rolling mills, so that its thickness is reduced to less than an inch. The problems in the condition and alignment of rolling mills can lead to significant quality issues. The most common causes may include the rolls failing to catch the metal, so that it will pile up or the rolls not rolling evenly, which results in one side of the metal sheet being thicker than the other. To prevent these issues, the condition and alignment of rolling mills’ bearings is monitored with vibration and magnetic guide sensors.

Automotive

In the automotive industry, the penetration of moisture into the spaces and gaps in welded spots can lead to porosity, while temperature variations in welding machines can lead to a weld joint failure. IIoT is applied to monitor the temperature and the level of humidity around a machine to avoid incorrect placement and ensure high quality of the welded products.

The benefits of product quality control based on IoT-driven condition monitoring

Monitoring the condition and the environment of machines, on which products are manufactured, delivers the following benefits:

• Compared to traditional production quality control techniques (for example, test checks carried out at the end of a production cycle), IoT-driven equipment condition monitoring lets users pinpoint quality issues at the production stage, when an issue can still be mitigated.
• The analytics capabilities of IoT-driven condition monitoring solutions lay a foundation for improvements in product quality. For example, combining historical data from vibration sensors attached to the milling rolls’ bearings with the data about past quality losses, manufacturers conclude that an 8% increase in a roller bearing’s vibration causes a metal sheet’s left side be 0.1 inch thicker than the right side. Manufacturers can then use these insights to improve the quality of the output products.

Getting started

Since the technology market does not yet offer out-of-the-box IoT-driven condition monitoring solutions, enterprises need to design and implement custom IoT applications. Given the complexity of IoT implementation, it proves efficient for the enterprises to collaborate with external parties: an IoT platform vendor or an independent IoT integrator.

Opting for an IoT platform vendor has the following advantages:

• Lower implementation cost;
• Simpler integration with enterprise and shop floor management systems;
• More comprehensive upgrades.

However, going for the collaboration with a single IoT platform vendor, enterprises are unlikely to get the best-of-breed functionality, as they often get locked up in the vendor’s solution ecosystem with limited options to test alternative solution components that may be a better fit.

Collaborating with an IoT integrator, on the other hand, offers the possibility to ‘build’ an IoT solution from the components tailored to the enterprise’s needs. Still, the cost of implementation will rise, as enterprises have to buy separate individual modules from multiple vendors and partner with an integrator to bring these modules together.

A point to consider

Although IoT-based condition monitoring paves the way to improvements in production quality, such an approach has certain limitations, as data about machine conditions may be not enough for well-rounded quality assurance. Monitoring the condition of machines, for instance, cannot identify issues arising from the use of defective or misidentified components, or improper material handling.

Controlling the quality of products by monitoring the condition of machines, on which they are manufactured, helps to drive yield improvement, reduce scrap, and minimize rework. Compared to other quality assurance techniques (for example, based on inspecting parts and semi-finished products as they move through the production cycle), the condition monitoring-based approach may offer less differentiation in terms of quality control scope, but it helps identify quality issues at their incipient stage and predict potential ones.