Part 2 in a 5-part series on “Design-to-Operate with the Digital Supply Chain.”
In this blog series, we’re looking at the design-to-operate (D2O) product lifecycle, which spans the phases of design, plan, manufacture, deliver, and operate. The first installment focused on design. Here, I want to shift the focus to planning.
Visibility and demand
The traditional supply chain planning function is undergoing a transformation. In the face of rising customer expectations around individualized products, shorter delivery timeframes, and predictable availability, organizations are moving away from the familiar cadences of tried and true planning approaches. Taking weeks to devise, propagate, validate, approve, and execute a plan is no longer an option.
Today, supply chain planning has to be continuous – happening in the here and now based on real-time information. With data regarding operations and the customer experience digitally available in real time, organizations can predict, plan, and drive goods through the supply chain more sustainably, profitably, and just-in-time. This helps organizations better cater to more demanding customers by offering more choices than ever before.
Planning at this level requires total supply chain visibility. Planners need real-time signaling that alerts them to what’s happening throughout the demand and supply network. Data on demand changes (real or predicted), manufacturing, inventory, logistics, and external trading partners needs to be factored into the continuously evolving plan. This can help drive more timely adjustments of long-term plans regarding new product launches, capital investments, or supply chain configurations.
The need to act and respond on a more real-time basis is driving manufacturers to get closer to the end consumer wherever possible. Manufacturers of consumer packaged goods, for example, may use social media and point-of-sale data to more accurately predict demand without the mediation of a retailer. High tech or consumer durables companies, meanwhile, may use IoT sensors to collect and analyze usage data. The result is an ability to not only predict demand but to directly impact the consumer experience and build brand relationships that last.
A new planning paradigm
In this world, the orientation is moving from supply-driven to demand-driven – with faster planning-to-fulfillment cycles that blur the lines between planning and execution. This new orientation, in turn, requires process and system change.
When it comes to planning material requirements or production capacity, for example, supply chain organizations are abandoning sequential batch processes in favor of integrated, self-regulating, and adaptive processes that focus on immunization against variability. At the same time, automation and technologies such as machine learning are leading to an increasingly “touchless planning” process. Today’s planners are empowered to focus more on solving problems and less on expediting processes.
Getting there with D2O
But how do you realize this vision? What’s needed is an integrated D2O approach that connects planning to all of the other phases of the product lifecycle – including design, manufacturing, delivery, and operations. Let’s have a quick look at how planning can intersect with each of these phases:
- Design: Typically, design is seen as the front end of the product lifecycle – disconnected from planning. In a D2O world, these phases converge – with planning providing critical and timely inputs on demand, production, and quality considerations right at the design stage. Planning, for example, can communicate projected volume targets and financial constraints that impact final design decisions. When supply networks are globally distributed, design can lean on planning to coordinate with contract manufacturers to plan and incorporate design changes at the right time and avoid excess and obsolescence.
- Manufacturing: In the face of fast-moving, highly volatile markets, manufacturing needs to be in lockstep with planning to keep in tune with the changing demand picture. When manufacturing encounters constraints that impact the plan, planning needs to adjust – for example, by aligning plans for re-staging raw materials to feed manufacturing and streamline capacity utilization.
- Delivery: Today’s manufacturers are moving aggressively toward same-day shipping models – where goods coming directly off the production line are picked, packed, and shipped in a matter of hours. Compressed cycles like these mean planning needs to be tightly integrated with logistics from the earliest stages. Help is needed to plan transport and monitor delivery for adherence to demanding service levels.
- Operations: Planning and operations are increasingly intertwined as the health and status of production assets and other capital equipment play a critical role in executing on any plan. For example, planning may want to track expected or unexpected maintenance closely to account for capacity bottlenecks, build ahead inventory, or switch to another line or facility. If the assets are used internally, this might drive future investment decisions and impact financial goals. If the assets are used at customer sites, the same information may indicate new service or sales opportunities.
Some help from machine learning
Across all of the interconnected phases of the D2O lifecycle, intelligent technologies play a key role in helping to streamline processes by providing insights that help speed up decision-making and avoid manual touchpoints. Let me end by pointing out just a few of the possibilities relevant to supply chain organizations:
- Demand sensing: Using machine learning, planners can detect patterns in how products are ordered and adjust the short-term forecast accordingly. The ability to understand how particular products are trending – in one direction or the other – can help you cater to demand and optimize production or transportation.
- Gradient boosting: Planners can also use machine learning for multivariable analysis against a wide range of data sets to uncover useful lead indicators. From weather patterns to competitor sales, analysis of this data can be used to substantially improve demand accuracy.
- Exception handling: With thousands or even millions of exceptions to contend with, planners can easily get overwhelmed. Machine learning algorithms can analyze these exceptions, check them for relevancy, and separate the signal from noise to prioritizes the most relevant.
The next blog in this series will focus on the manufacturing phase of the D2O lifecycle. Be sure to stay tuned.
In the meantime, if you need more information on D2O, have a look at the new whitepaper from IDC Planing Steers the Digital Supply Chain.