The Killer App for IoT in the Supply Chain

Remember RFID?

It wasn’t that long ago that RFID was touted as the solution to solve a whole host of supply chain problems. Improve supply chain visibility, reduce operational costs, reduce inventory and even make a mean Martini! Manufacturers of consumer goods such as P&G and retailers like Walmart were early and vocal proponents. To be fair, RFID did help several supply chain organizations streamline operations and improve inventory tracking - especially in the areas of retail, warehousing and transportation. And outside the world of supply chains, RFID, in some form or the other, has permeated our lives (think toll tags and passports). But the jury is still out on whether RFID lived up to its promise, or at least the promise that was sold to the broader supply chain community.

These days there is a new kid on the block – The Internet of Things or IoT for short. If it feels like Déjà vu all over again, you are not alone. Surrounded by catch-all terms like ‘The Digital Enterprise’ or ‘Industry 4.0’, and mixed in a word salad along with Cloud, Big Data and Machine Learning, it is difficult to separate the hype from reality. No wonder that supply chain leaders have yet to get on board the IoT train.

Let’s take a classic use case for IOT – predictive maintenance – and use it as an example to dig deeper and understand the promise of IoT. And as a bonus, come up with what might just be the killer app for IoT in the supply chain.

 A tale of two machines

Say you run a factory that makes widgets. Half the widgets are made on Machine A, an old but still reliable workhorse. The other half are made on Machine B, a shiny new version of machine A, but with the same average production rate.

Machine B has advanced sensors that capture every detail about its operations – hours of operation, number of cycles, load conditions and so on. It transmits that in real time to the maintenance department who use predictive maintenance software to determine maintenance requirements using reliability parameters and sensor data.

• You now have a much better handle on how much maintenance spares you need to stock. You also proactively replace parts just before they are most likely to fail thereby reducing downtime. You save money by not replacing the parts that don’t need replacing based on the machines specific operating conditions.
• Not happy yet? The predictive maintenance software also predicts the likelihood of downtime which is fed into the planning and scheduling system used by the factory. This allows you to plan ahead of likely downtime, reducing the need for safety stocks of finished widgets to protect against unknowns and still keep customer service levels up.
• But wait, it gets even better - The sensor data is also fed back to the machine manufacturer in near real time. They in turn use this to generate a more accurate driver-based forecast of spare parts that reduces their stocking levels.
• And to top it all - sensor data from many such machines in the installed base is combined to help the machine manufacturer’s design engineers to improve product reliability even while rationalizing design safety factors – making future machines lighter, cheaper and more efficient.

A virtuous circle and a win-win for all. Machine A on the other hand, produced the same average number of widgets but could not claim these benefits. Why? What was happening to Machine A during this time? You maintained more just-in-case maintenance spares. You focused on reactive maintenance causing unpredictable downtime. This means you kept additional safety stock of finished widgets. The machine manufacturers continued to be reactive and keep additional spare parts stocks. Finally, there was no feedback loop to improve product design, so they continued to pad with higher safety factors.

 Looking under the hood

While this was a specific use case for predictive maintenance, an observant reader would have noticed a larger pattern. Each of the listed benefits of IoT based Predictive Maintenance was directly linked to reducing variability of a process and by extension, reducing the buffer (inventory for example) needed to support it.

That, right there, is the killer app for IoT in the supply chain - The ability to measure the underlying (previously hidden) variables of any process, then improving the predictability of that process and therefore reducing the buffer needed to maintain the same level of throughput.

This implies that any number of supply chain processes could potentially benefit from IoT; with the prime candidates being areas where

1. There is high variability inherent in the process (variability of the output is not mainly due to variability of the input).
2. The buffer needed to ‘manage’ that high variability is costly or infeasible.

Closing Thoughts

IoT is not just old RFID in a new bottle. Along with other co-evolving new technologies in analytics and machine learning, it has the capability to revolutionize industry as we know it. But we need to go beyond the hype and really understand at a fundamental level what IoT has to offer the supply chain.

When faced with a technology landscape that is changing so rapidly that you question your ability to keep up, it might help to take a step back. And to paraphrase John F Kennedy’s speech, instead of asking “What can I do with IoT in my supply chain?”, we should be asking the question “What do I need to get done and how can technologies like IoT help me?” The two questions sound similar, but are subtly different. Instead of putting the technology (IoT) at the center as in the first question, in the second approach we turn around and put the business first.

Seen from this perspective, IoT is a technology enabler of business processes whose underlying fundamentals have not changed significantly. Replacing inventory with information. Imagine that!

Disclaimer: This publication does not represent the thoughts or opinions of my employer. It is solely based on my personal views and as such, should not be a substitute for professional advice.