Inside Logistics

Is Just-in-Time out of time?

JIT is built on the assumption of stability, but in a world of volatility, logisticians can no longer afford to assume stability


June 2, 2020
by Paul D. Larson

Larson

Paul D. Larson, Ph.D., is the CN Professor of Supply Chain Management at the University of Manitoba.

On a chilly, sunny April morning in Winnipeg, the author was sent to a nearby store to buy toilet paper, just-in-time. The aisle dedicated to paper products was barren. Such a stunning, sudden logistics service breakdown!

Logistics and JIT

Stated most simply, logistics is the management of inventory – in motion and at rest – to serve customers. Transportation and warehousing are the primary logistics functions involved in the motion and rest part of the equation. Logistics managers balance customer service and total costs. The beauty of just-in-time (JIT) is its promise to maintain service with lower inventory levels – and lower costs.

But JIT is built on the assumption of stability, i.e. level demand and short, reliable lead times. Sudden supply chain disruptions, such as fires, floods or pandemics, bring instability. Demand surges for certain essential items – and plunges for others. Upstream shortages mean longer, less reliable lead times.

Forward buying – sometimes called panic buying – means purchasing volumes in excess of typical demand. It is triggered by anticipation of impending shortages or rising prices. Erratic demand and unreliable delivery invalidate the assumptions enabling JIT.

It’s elementary!

Prominent among the promises of JIT is reduction of three types of inventory: cycle stock, pipeline stock and safety stock. Given stable demand, shipment quantity (Q) can be set to minimize the sum of ordering, transportation and inventory carrying costs. If daily demand (r) is constant, which is the ultimate in stability; cycle stock ranges from 0 to Q units and averages Q/2 units. A shipment arrives, it is used at a constant rate, and the next shipment arrives just as the previous one runs out.

The only way to reduce cycle stock is to reduce Q. Small shipments or lot sizes is a well-known feature of JIT. But the further away your supplier is, the more likely the economics of long distance transportation dictate large shipments, e.g. full containers. JIT works best with stable demand and nearby suppliers. A complex, long, global supply chain – and the large shipments it impels – compromises the JIT quest for low cycle stock.

Pipeline stock is inventory somewhere on the way between origin and destination, between supplier and buyer. It is often moving on freight vessels or vehicles, such as planes, trains, trucks or ships. It might also be moving through a warehouse or waiting at a national border, intermodal freight terminal or receiving dock. Q/r is the number of days a shipment of Q items will satisfy daily demand. Its reciprocal, r/Q, is the average number of shipments dispatched per day. (If Q = 5,000 and r = 1,000; then a shipment is dispatched every fifth day. On the other hand, if Q= 1,000 and r = 5,000; then five smaller shipments are dispatched every day.)

Lead time (t) is the number of days a shipment is in motion or at rest between origin and destination. In a stable system, shipments of size Q are dispatched at a rate of r/Q shipments per day and spend t days moving or waiting to move between origin and destination. Pipeline stock = t * Q * r/Q = r * t, i.e. on average there are r/Q shipments of Q units per shipment in the pipeline for t days. The only two ways to reduce pipeline stock are to reduce r (rarely if ever desirable) or to reduce t. Once again, complex, long, global supply chains strain JIT objectives, by making lead time reduction more challenging.

The third type of inventory, safety stock, is needed if demand or supply variability is expected. Under conditions of variability, demand is often assumed to follow a normal distribution, and safety stock is set at: service level factor (z) * standard deviation of daily demand (sr) * square root of lead time (t)1/2. Based on the standard normal probability table, for a 95 percent service level (e.g. 95% of orders filled same day from stock), z = 1.645. A higher desired service level would mean larger z and more safety stock. More variable demand and longer lead times also require more safety stock to maintain the desired service level. So, the only ways to trim safety stock are to reduce service level expectations, make demand less variable, or reduce lead time. As noted above, lead time usually increases with distance between supplier and buyer.

Disruptions, like the current pandemic, also increase lead times; while ramping up variability of demand (surging for essential items, plunging for non-essentials) and heightening service level requirements for critical items, such as personal protective equipment (PPE) or TP.

What’s a logistician to do?

First, reduce upstream lead time and lead time variability. This might mean moving to multiple sourcing and/or local sourcing. During a disruption, distant, single sources are risky in terms of lead time. Local sources are likely able to respond more quickly.

Second, anticipate volatility. Of course, some disasters cannot be anticipated. What about the coronavirus pandemic? In late December 2019, a physician in Wuhan, China posted warnings about a SARS-like virus. On January 30, the World Health Organization (WHO) declared a global public health emergency. The very next day, the U.S. blocked foreign nationals who had been in China within the previous two weeks from entering. How many supply chain managers in JIT operations anticipated the coming volatility in both lead time and demand? How many placed larger than normal orders for critical inventory items in January?

Moving forward, logistics managers are advised to partition their inventory items, considering importance and volatility of supply and/or demand. Often item importance is based on annual volume X dollar value. So, a million dollar item with volume of one is just as important as a $1.00 item with a one million unit volume. However, there is more to item importance than volume X value. If an absence of light bulbs or protective gloves shuts the operation down or endangers employees, those items are important, irrespective of volume and value. If a stock-out means losing customers (or employees) forever, the item is important. Logisticians must ask: What are the costs and benefits of maintaining inventory?

If a disaster is anticipated, logisticians should build safety stock equal to at least demand during lead time for the important items. So, if expected lead time is 50 days and daily demand is 100 units, hold 5,000 units in stock. These items could be warehoused in a centralized location, as long as fast transport is available when needed.

It’s about time

Is JIT out of time? No, but it is about time! JIT is making and moving things just as needed. It has always been about time – and it still is. It’s about reducing lead time, but also about under-standing determinants of lead time. JIT is built on the assumption of stability, i.e. level demand (r) and short, reliable lead times (t). But in a world of volatility, logisticians can no longer afford to assume stability.

Today, in Winnipeg, there is an ample supply of toilet paper on store shelves once again – and a two-month supply in the author’s basement, just-in-case.