How the internet of things can revolutionize product correction cycles
Internet of things (IoT)-equipped products have the potential to hugely reduce the cost and time involved in detecting and correcting faulty products.
When most people talk about the IoT, the immediate connotations are of the consumer side of the technology: the futuristic, high-tech houses where appliances speak to each other and anticipate their owner’s every need. But, the impact of the IoT will likely first be felt in the industrial sector, where smart machines and data analytics could have a huge impact on a wide range of traditional manufacturing, infrastructural and maintenance projects. One area in which IoT solutions can potentially add a great deal of value to the manufacturing process is in error detection and correction.
Warranties versus data
Today, when an error in a product occurs, the usual means of managing it is through warranty submission analysis. Let’s use an example to illustrate the process and structure. Say there is a fault with a car, picked up by a consumer who submits a warranty to the manufacturer. The next step is that the vehicle is submitted for inspection which creates data that can be analyzed and action taken to deal with the issue. Part of this resulting action could be a product recall (which one North American automotive manufacturer indicated can equate to US$1m a day in warranty and recall related costs). The entire process of warranty-based correction can take between 120 and 200 days.
Now, let’s say an IoT-fitted, connected car breaks down. At the moment of breakdown, the sensors in the car create what is known as a diagnostic trouble code (DTC), a data set capturing the exact digital status of the vehicle at the moment of failure. This data is then transmitted to the manufacturer, where it is analyzed and actioned. IoT solutions such as this can cut up to 90% off the detection-to-correction (D2C) times, down from hundreds of days to just a few weeks, or even a few hours. IoT solutions can also capture data on device conditions that a conventional car wouldn’t, such as driver behavior, leading to more accurate and valuable diagnostics. Overall, IoT solutions could save manufacturers about US$1.8b per incident (see diagram).
The data game
These IoT- and analytics-driven solutions depend on a complex architecture of supporting data systems and methodologies. The first issue is simply storage — the average IoT-enabled vehicle generates around a petabyte of data every 6 to 10 hours it’s in operation (for scale, a petabyte is about four times as much information as the US Library of Congress has collected in its entire history). This is stored on giant server farms, on top of which sits an analytics framework that allows for the data to be processed.
But even when this data is structurally supported, there’s still a huge amount of material to sift through to look for useful data points. One method of simplifying this task is through product reliability analysis, which can determine the rate at which a certain product fails, quantified as the mean time between system failures. This rate is called the “accelerated failure time” (AFT), and can let engineers identify factors that contribute to faster failure rates. This could include speed of moving parts wearing down metal, or the quality of the materials themselves.
Lying on top of this is a final analytics layer called the RAAK algorithm. Engineers can use the RAAK algorithm to aggregate AFT metrics together. The algorithm then produces a chart that depicts the AFT data as a series of bubbles sized in relation to the relative potential severity of the factor each represents. These visuals can be modified in real time to provide an ongoing representation of the development of risks as they develop, in an easily consumable form so that a broad range of stakeholders can take action where appropriate.
The key value of IoT in error detection and correction and maintenance solutions is in the speed with which they can roll out. Now, teams can know when, where and how a breakdown happens almost as it occurs, and deliver solutions accordingly, while management can take a broader strategic approach to product design and development. It’s in these behind-the-scenes roles that IoT is able to change dramatically the way organizations plan and work.