COVID-19 taught us the importance of supply chains when everything from raw materials to finished goods became delayed or simply unavailable to manufacturers and retailers. It also accelerated a dramatic shift in the logistics and delivery side of the supply chain equation as consumers moved from brick-and-mortar purchases to online shopping. The dynamic nature of the full supply chain is now a given, demanding significant shifts in the way we view optimization.
The goal of a supply chain organization is to meet customer requirements while minimizing total supply chain costs. Businesses must be flexible enough to respond quickly when disruptions occur.
Unfortunately, most of us aren’t as agile as we could be, as this research from Ventana points out:
- 79% of companies use spreadsheets for supply chain planning.
- Less than 25% say their supply chain plans are integrated with their company’s manufacturing, procurement, or sales departments.
- 54% say they have limited or no ability to measure supply chain trade offs across departments when making decisions.
Additionally, the last mile grows even more complex. The last mile has always been the most expensive, long-bemoaned challenge of the supply chain. With the “new normal” of changing consumption habits and channels creating unpredictable demand, forecasts have become meaningless. This makes agility and speed to optimization that much more important to meet customers’ growing expectations for instant availability and near-immediate delivery.
A fixed logistics model is not designed to be flexible or fast. Capgemini Research Institute, Supply Chain Survey 2020 found that 70% of companies are prioritizing inbound and outbound logistics as part of their supply chain sustainability efforts post Covid. Yet less than half of organizations asked by Accenture agree they’re currently meeting customer expectations for order fulfillment.
What happens when the industry becomes even more dynamic and customer expectations require that time cycles compress?
The Missing Link
Constrained optimization helps manufacturing supply chains by identifying the best path forward as dynamic conditions impact sourcing and logistics options. In simple terms, constrained optimization guides you to decide how to do more with less, or how to use less to do more.
Most economic business decisions require applying constraints, such as cost, volume, or time to a set of variables, such as trucks, SKUs, or people with an objective to minimize (cost) or maximize (profit) outcomes. Every organization has a multitude of such optimization problems to solve.
This sounds like something we should be using, right? But there are a few reasons why we don’t:
- Enterprises with major investments in big data and analytics may assume they have the analysis and reporting in place that’s needed. But classical computing can’t process the volumes of data we’re collecting. Analysts end up compressing and reducing the data to run a computation thereby reducing the data analyzed to get a result.
- Information is analyzed and presented to management and decision makers. They apply their unique perspectives to the debate and decide — the way they’ve always done it. And they assume that’s enough. Constrained optimization takes the debate and personal filters out of the equation to show you the best decisions.
- Data isn’t just data. It has interrelationships with other data you must consider to get accurate and high-quality answers to optimization requests. If a classical computer doesn’t completely falter, it may only give a single probable answer, and it may or may not be accurate.
Classical vs. Complex
Many of us have heard of the traveling salesman problem, which can be compared to truck routing and how to optimize the routes, as well as the trucks. The challenge is that traveling salesman problems like this grow in complexity by n! (n factorial). Routing problems are more constrained and complex for every variable (truck, route, driver, etc.) that you add. For example, a traveling salesman problem that has 10 stops results in 3,628,800 route options, 40 stops will result in approximately 40! = 815,915,283,2 00,000,000,000,000,000,000,000,000,000,000,000,000 options. Routing multiple trucks and packages is even more complex.
A classical computer would struggle under the weight and scale of a vast set of possibilities. This is where quantum computers promise to take on the task to quickly produce options to choose from to make the best decision based on your goals.
Complicated scenarios meant to solve for multiple variables are not achievable by a classical computing algorithm in a short span of time. However, algorithms using quantum computing techniques can quickly achieve this simulation using a classical system applying quantum techniques, or a hybrid solution that employs both quantum and classical, today.
Accenture concurs, stating, “Route-optimization algorithms are helping reduce mileage and improve on-time delivery rates. In logistics, quantum routing uses cloud-based, quantum computing to calculate the fastest route for all vehicles, taking into account millions of real-time data points about traffic congestion.”
Here are a few additional ways constrained optimization benefits manufacturing supply chains, from inbound raw materials to outbound distribution:
- Transportation efficiency. Constrained optimization is used to identify optimal locations for plants, distribution facilities, and other logistics hubs. Even a mile difference in placing a plant can make a significant difference in the costs and productivity of the overall network.
- Warehouse management and distribution. Constrained optimization is applied to optimize global and local shipping and loads, warehousing and delivery for lowest cost, optimum efficiency, and productivity. Imagine having to schedule shipments of thousands upon thousands of computers, televisions, or cars across the globe using a piece of paper or a spreadsheet.
- Inbound logistics. From order levels to delivery to the production line, optimization can drive maximum production levels at the best cost. Even one lost shipment or forgotten vendor can wreak havoc on a production line. Now imagine having to schedule and maintain all the parts of a car, computer, TV, refrigerator, or ATV using a spreadsheet, especially when you’re managing hundreds of thousands of units.
IDC research concludes: “The ability to ingest broad and deep data sets to inform better decision making will be the single largest differentiator of supply chain performance in the future.” Quantum computing techniques empower constrained optimization to a new level of accuracy and performance.
Quantum computers process complex computations to return a diversity of answers, not just one. Every answer that meets the optimized state you need is delivered to you. You get exposure to more viable options than with classical processors and can select the one that best matches your specific situation right now. This is a much better way to make decisions vs the classical software approaches that provide a single answer as your only option.
Quantum computing is one of the most promising technological innovations likely to shape, streamline and optimize the future of the supply chain. It offers better insights to make better decisions. That’s why there’s so much excitement about it.
Robert Liscouski is president and CEO of Quantum Computing Inc.