Let IoT Analytics Breathe Life Into Your Supply Chain

COT WP
Whether you’re a local flower shop, a Fifth Avenue furrier or a Fortune 500 powerhouse, your livelihood depends on the effectiveness of your particular supply chain. Think of the supply chain as a living organism: some larger, some smaller, but all essential to the wellbeing of the business.

Think, too, of the advent of Internet of Things analytics as a kind of Fitbit for giving you visibility into your supply chain’s health and performance potential. IoT analytics are a natural fit because they can instrument and report on nearly every movement and handoff that goes into getting your source materials to your door.

In fact, supply-chain operations are so important that business has for years been dedicating ERP and other resources to their continuous improvement.

But something is different today, and it comes with the ability of new cloud-enabled IoT architectures to deliver analysis, and thus end-to-end visibility, in real time.

Of course, getting data in real time isn’t by itself a major breakthrough – a phone call from a supplier saying that a shipment will be delayed is by definition a real-time event. But having the technology framework in place to analyze that data, now coming in by the multi-petabyte, is a breakthrough.

New: The Always-On Supply Chain

Thanks to the IoT, the entire supply-chain information model is now changing from a checkpoint-to-checkpoint reporting mechanism to an always-on producer of continuous events, from ecosystem suppliers and manufacturers to transporters, warehouses and distribution centers.

IoT analytics take that detail and merge it upward to create a new plane of visibility, where CEOs, CIOs, CFOs, manufacturing and logistics managers, business unit executives, planners and other authorized users can drill down from big-picture dashboards into minute detail at the sweep of a finger.

For example, by combining weather information with GPS location data a logistics manager might decide to reroute a shipment of time-critical inventory. Or by analyzing a supplier’s production output a manager might decide to consolidate shipment of the newly manufactured parts with the output of another supplier located close by.

In another example, a customer-service manager might suddenly spot a slowdown in a scheduled customer delivery. The service manager can then text or call the customer with a proactive update…rather than waiting for the customer to call with a complaint.

Other examples, as well as further examples of analytics technologies, are available in the KPMG publication “Supply Chain Big Data Series: How Big Data is Shaping the Supply Chains of Tomorrow”.

How Supply Chain Analytics Work

At a high level, an analytics engine takes in real-time data coming from sensors, beacons and other sources across the supply chain. Signals can come from suppliers’ factories, from warehouse cameras and sensors, from ships and delivery trucks, from GPS telematics, from voice audio or call logs, and from the networks that service those sensors.

Operating at ultra high speeds, the analytics engine is also searching data stores and other local sources for relevant historical and situational information. If a warehouse delivery is taking longer than expected, for instance, the historical database might show that this is a recurring pattern for this particular warehouse at this particular time of day.

The analytics engine will also access relevant machine-learning or other models to predict likely outcomes, and even to prescribe specific actions, like quickly switching to a secondary warehouse. And all of this activity, taking place in microseconds, is showing up on the dashboards of company managers and planners.

The result is a plane of visibility that delivers end-to-end views of the entire process, views that no single supplier could see. And it lets the operator take immediate corrective action. More advanced analytics engines can also trigger immediate process workflows to work around the problem, or to take advantage of an unusual situation.

Benefits include an overall improvement in supply-chain agility, and filter down to a variety of competitive advantages in cost cutting, product pricing and customer service. Also, greater benefits accrue over time as the analytics architecture takes in more and better contextual information and updates its models accordingly.

Planning Starts Now

With IoT technology becoming more prevalent by the day, the possibilities for future uses – and benefits – seem limitless. And in a sense they are, since it’s difficult today to predict the kinds of models that will be available in just a few years’ time. But along with these seemingly limitless use cases comes a significant real-world challenge: how best to go about building the analytics infrastructure that can function above, or within, the complexities of today’s network and IT infrastructures?

There’s no simple answer, of course. But certain fundamentals will apply regardless of an organization’s industry type or current technology architecture, even if it consists of mixed on-premise and cloud systems.

For fundamentals, it’s best to start planning for:

  • An open-ended, cloud-based IoT analysis framework – So-called point analytics solutions can be effective, but over time they run the risk of adding to the complexity of enterprise IT operations as business unit managers bring in varieties of specialized applications. An end-to-end framework creates an orderly context for applications. It maximizes corporate flexibility for moving quickly into new analytics pursuits as strategic priorities change, and it minimizes the costs and risks of trying to govern across overlapping analytics initiatives.
  • A production-quality app development environment – Because so much is still unknown about the needs and direction of future analytics applications, it’s essential to give data scientists and business analysts the best possible tools for creating those new apps. The development environment should be modular to maximize reusability of components; it should be open to current and future analytic programming languages; and it should employ graphical techniques that facilitate communication and understanding between developers and business analysts.

The good news is that current cloud, visualization, machine-language and other technologies are available today that deliver these fundamentals, and can help organizations get started immediately on their analytics journeys.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>