A Tale of Two Wines

I’ve spent most my life in the Bay area. The weather is mild, and the traffic not. Over time, I’ve become fond of a particular French wine sold in a chain grocery store. The wine had good character and depth, but its price initially made me think I’d be drinking vinegar! So, when I had to relocate to (hot! hot!) Arizona for two years, I found the same chain grocery store to buy my wine. Everything inside the store was arranged in exactly the same way as in California.  More surprising, they had all the same products, including perishables, as in my California store. And there it was -- my favorite cheap wine! I thanked the supply chain gods for making this happen.

I took my wine to the corporate housing, feeling very much like I was home. I opened the bottle and let the wine breathe. First sip… it tasted not just like vinegar, but bad vinegar! Second sip… still the same. I went back and got another bottle… same taste. Obviously, somewhere during the transit of the cargo, this delicate, cheap wine was exposed to the hostile Arizona heat for too long. That killed it. I was so disappointed.

So what does this really have to do with the supply chain process you ask?

Decayed Links in the Supply Chain Process

When we started doing IoT development at Oracle, one of the most common customer concerns was reducing theft and waste in the transport of perishable goods (for example, groceries, poultry) and semi-perishable goods (such as wine). Though we offer the best practices to tune your supply chain logistics to reduce your hazards when moving your perishable goods, we realized many risks still weren’t being  addressed.

Risk 1: Asset Is Missing

What do we mean by assets? An asset may be a keg waiting to be loaded in a yard, or it could be an excavator, a loader, lathe, drill, blood infusion pump, or a patient bed. Assets could be goods moving through the supply chain system, or a tool enabling the movement of those goods. Some of us sheltered within our software world don’t ever consider the loss of physical assets – it takes us by surprise. How can an asset get lost?

Consider a yard in a typical manufacturing plant. Unlike digital records neatly residing in database tables, no dictionaries document the layout of this yard. From a layperson’s view, sometimes it just feels like there are disorganized scrap piles all over the place. It could be only a matter of time before someone puts something in the wrong pile. Even when the asset isn’t technically lost, it takes valuable time to locate that asset, corroding the productivity within your supply chain. This problem of missing assets can happen across the board – in manufacturing plants, breweries, hospitals, conferences, construction yards, and so many other businesses.

So what’s the solution? Geofences – that is, set up a virtual yard whose geofence boundaries help you track your assets. Using geofences, you can track your assets not only between organizations, but also even within your organization. You can track all of your many assets, and get alerts in real time if any asset crosses the borders of your digital fences.

Risk 2: Asset Can’t Perform

Not all assets work all the time – this is just a fact of life. For example, think about the last time that you rented a car. The car rental service, as the service provider, schedules enough time between rentals to clean, service, and repair the car. But sometimes this isn’t possible. For example, in the last shift, the car rental agent forgot to turn the lights off on the car that you turned in. As a result, in the morning there’s one car that can’t be rented because its battery died and has to be recharged or replaced. Your car rental business won’t be as productive today until that happens.

And because this is an unscheduled maintenance, there aren’t any mechanics immediately available to fix the car, and they’d have to reschedule their existing appointments. This requires filing a service ticket. Your supply chain is losing valuable time during which this car could have been rented and earning some revenue.

So what’s the solution? Performance and condition tracking – that is, track the performance and the condition of any asset in real time, at any time.  We expanded the tracking feature to include business foresight: Make timely predictions about when your assets need scheduled maintenance. Our IoT supply chain process deeply instruments your asset, which means that you can order replacement parts as needed based on scheduled service. Best of all, you can automate the creation of a service ticket for your asset – no excess, time-consuming paperwork!

Risk 3: Asset Is Stale

Ever taken your iPhone to the Apple Store to get it serviced? I have. One of the first things that they check after opening the case is a sensor that tells if the phone has been exposed to liquids (I accidentally dropped my phone in my morning tea). If the sensor reveals that the phone has been exposed to fluids, then the warranty is negated. I’m happy to say that they still gave me a free replacement because they loved my sob story.

This risk is similar to environmental monitoring of the fresh or stale status of an asset. The more sensitive an asset (that is, the greater the risk of becoming stale), then the higher the fidelity that you’ll need to provide end-to-end monitoring in the supply chain life cycle of an asset. Remember my wine? The temperature-regulated journey from France to Arizona may have taken weeks, but within just a few hours of the wine being left in, well, Arizona weather, it became vinegar. This is important to note: Yes, today we have refrigerated cargo containers; sometimes we even track them in real time. However, tracking a cargo container doesn’t guarantee that its cargo has been in the desired environment throughout the entire supply chain cycle.

What we need to track is the wine, not its cargo container.

So what’s the solution? At Oracle, we have various implementations deployed to monitor all kinds of goods such as groceries, beer, young livestock, and so on. What we’ve learned is that environmental monitoring isn’t as easy as implementing threshold alerts (for example, temperature is > 90 F). The metric that’s developed is usually a function of readings from multiple sensors and other context that may be residing in back-end information management systems. For example, in one of our implementations where we were transporting newly born livestock, we developed a metric based on ventilation (sensor 1), temperature (sensor 2) and age of the livestock (information from the enterprise system).

Many times, we don’t immediately know if an asset has become stale in real time. For example, there could be extenuating environmental conditions that we still haven’t identified that could be causing premature staleness. For that reason, it’s vital that the environmental conditions persist. Later, during the supply chain process, if there are deviations in product behavior, then we now have the ability to derive new insights using advanced analytical techniques, including machine learning. For example, we may discover that the wine from a particular vineyard is more sensitive to temperature, so it needs to be transported at lower temperature levels.  

Risk 4: Asset Health Disconnected from Business Processes

One of the mistakes that I’ve seen early practitioners of enterprise IoT make is building isolated systems that focus exclusively on the physical and operational parts of the supply chain/IoT system. By doing so, they miss the opportunity to get the benefits from automating compliance and work orders.

Let’s look at automating the work order. It isn’t as simple as tracking an environmental condition and triggering a work order in your field service application. Sure, this is useful if a product has broken down and we’ve determined that a work order needs to be created. However, this is a low-value automation. Often, we pick up signals that only suggest that something requires attention. More useful examples of key performance indicators (KPIs) answer the following questions, in real time:

· Is my asset in the bottom 10% of its historical performance?

· Has my machine been serviced as prescribed in the schedule?

· Are my alerts coming from a particular shift/operator/factory?

· Is my asset performance seasonal?

Depending on these metrics, different corrective actions need to be taken. To develop such forecasts, the system needs to have a tight coupling with your back-end systems where asset maintenance and productivity information is stored. In addition, your system requires powerful analytical capabilities to detect outliers and trends. By capturing environmental and asset conditions throughout the entire supply chain life cycle, we can automate the gathering of information about how well your process complies with the previous KPIs.

So what’s the solution? Enterprise IoT asset tracking – all of the previously mentioned risks exist in all businesses where enterprise IoT is being adopted. In developing our asset tracking application, we set these goals that all our customers should achieve with our product:

1. Extend asset life to avoid expensive capital purchases.

2. Track all assets per your own guidelines.

3. Identify personnel or equipment investments to maintain service levels.

4. Improve asset utilization.

We also understood from our experience that IoT adoption for asset tracking isn’t a one-step process. The first step is to integrate asset tracking with your existing asset management system and then visualize it in a map-based dashboard. Next, you enable various sensors and set up geofence rules to detect if any of your assets has been misplaced. After these steps are completed, you can define rules to track the condition of your assets and integrate this information with your work order management system. As soon as these steps are complete and operational, your implementation team regularly reviews the historical data with advanced analytical tools to reveal new business insights.

In our coming blogs, we’ll cover the design principles and IOT Asset Tracking features in more detail.