scheduling – The OM Blog by Heizer, Render, & Munson

Guest Post: How AI Has Transformed Semiconductor Scheduling at NVIDIA and TSMC

Professor Misty Blessley raises an interesting AI issue-chip plant scheduling.

Semiconductors have been called the brains of the modern world. Switching between conducting and blocking electricity, they are essentially “on/off” switches that control the flow of power, and this unique characteristic makes them the building blocks of modern technology. They are behind computing, communications, and energy innovation, and our information dependent society is always hungry for more. Leading chip makers, NVIDIA and TSMC, both use an AI to advance semiconductor design, scheduling, and manufacturing, report ElectronicsUSA (June 4, 2026) and NVIDIA News (May 31, 2026).

TSMC ( Taiwan Semiconductor Manufacturing Company), the world’s largest semiconductor foundry, makes the most advanced chips on the planet. NVIDIA is a global leader in accelerated computing and AI. For decades, TSMC has manufactured NVIDIA’s chips, and this partnership has come full circle. TSMC uses NVIDIA’s AI technologies inside its fabrication plants (“fabs”), which in the semiconductor industry refers to highly specialized facilities where silicon wafers are processed into microchips.

NVIDIA’s AI models are now embedded directly into TSMC’s manufacturing workflow, transforming production scheduling. A single wafer, a subcomponent of a semiconductor, may require hundreds of tools and thousands of tightly sequenced steps. TSMC can now evaluate millions of scheduling combinations in seconds. This optimizes job sequencing, and instantly rebalancing schedules when tools go down or urgent orders arrive. The result is smoother production flows, less idle time, and higher overall fab productivity. AI comes as close as ever to “running the plant.”

AI driven scheduling is also advancing through emerging capabilities such as predictive dispatching, where models forecast bottlenecks hours ahead and reroute wafers to prevent delays. Another is crossfab load balancing, which evaluates capacity across multiple TSMC sites and shifts work to maximize throughput. Both approaches are expected to reduce fab cycle time 5–10%.

NVIDIA’s AI scheduling helped TSMC cut critical chip production workloads 20-50%. This addresses society’s insatiable demand for semiconductors. It also fuels the next wave of breakthroughs in computing, communications, and energy innovation, the domains powered by these tiny on/off switches.

Classroom discussion questions:

What are the strategic and operational risks if a company relies too heavily on AI to automatically reroute work without human oversight?
Discuss how AI’s ability to evaluate millions of combinations in seconds changes a manager’s approach to bottleneck scheduling compared to manual or sequential sequencing rules (like SPT or FCFS).

OM in the News: Oregon’s Basketball Scheduling Nightmare

operations management and NCAA and basketball When the University of Oregon announced last summer that it was joining the Big Ten conference, it discovered that its team would spend more time this season up in the air than actually playing basketball. Since their season began in November, the Ducks have crisscrossed the country so frequently that the total distance they’ve traveled this season amounts to 26,700 miles, the equivalent of traveling the entire circumference of planet Earth.

Oregon is by no means the only team racking up air miles this season, reports The Wall Street Journal (March 15-16, 2025). The most recent wave of conference realignment has stretched the bounds of geographic imagination in college sports, sending schools like California and Stanford to the Atlantic Coast Conference and plunking UCF (in Orlando) in the Big 12, traditionally based in the Great Plains.

The Big Ten did its best to design a schedule that mitigated the impact on West Coast teams. The conference tried to avoid scheduling away games when players would be taking final exams and sequenced opponents so that traveling teams could play two games on a single road trip. “The scheduling of it,” said the Big Ten CEO, “is a sport in itself.”

Doing this in practice, however, often meant Oregon was forced to spend several days away from campus. When the Ducks made their first cross-country trip to play Ohio State and Penn State in January, the team spent 6 days and 5 nights on the road. More troubling than the flying was the amount of time-zone hopping the team was forced into, which included two road trips to the eastern time zone and two more to the central. Internal clocks are most thrown off by eastward travel of 3-5 hours, which can negatively affect reaction time, concentration and athletic performance.

That’s not the only reason that Oregon’s road trips are more disruptive. In the Pac-12, league games only took place on Thursdays and Sundays. The Big Ten plays every night of the week. To make all the travel work with academics, every player at Oregon had enroll in online classes this term—the first time that’s happened.

The graphic below shows intraconference away game travel for former Pac-12 teams. 2023-24 is on the left and 2024-25 is on the right.

Classroom discussion questions:

How can OM help with this problem?
What about other college sports?

Guest Post: Quality, Marketing and Cross Contamination

Professor Howard Weiss shares his thoughts about a variety of unusual OM topics with us monthly.

People with food allergies typically check the ingredients of a food product very carefully to ensure that the product does not contain an ingredient to which they are allergic. The top 8 allergens in food production are soy, wheat, milk, peanuts, tree nuts, fish, and shellfish.

Bimbo Bakeries, headquartered in Mexico with bakeries in 35 countries including the U.S., has taken a unique approach to listing allergens on some of its products. U.S. inspectors reported that Bimbo Bakeries USA — which includes brands such as Sara Lee, Oroweat, Thomas, Entenmanns and Ball Park buns and rolls —”listed ingredients such as sesame or tree nuts on labels even when they weren’t in the foods.” (Bimbo claims to be the largest bakery in this country).

The reason a company might purposely list ingredients that are not in its products is that it may be concerned about cross-contamination in a bakery plant and wants to ensure it will not be legally responsible in the event of cross contamination. In other words, rather than trying to introduce quality control procedures to prevent cross-contamination in its plant, the company is willing to be untruthful when listing ingredients to minimize the chance and or cost of a law suit.

It may be very expensive or difficult to prevent cross-contamination from one part of plant to another or from one machine to another. So to stay within the letter of the law some companies have deliberately added small amounts of allergens to products that previously did not contain these allergens. This helps the company avoid liability and legal costs.

Cross contamination can occur in several different ways:
 primary food production — from plants and animals on farms
 during harvest or slaughter
 secondary food production — including food processing and manufacturing
 transportation of food
 storage of food
 distribution of food — grocery stores, farmer’s markets, and more
 food preparation and serving — at home, restaurants, and other foodservice operations

There are strategies available to minimize the chance of cross-contamination. The best way is for food manufacturers to process products that contain allergens in a separate facility. If this is not possible then scheduling the production of products that contain allergens at a different time than other products may help. Cleaning procedures can be used to minimize the chance of cross contamination.

OM in the News: The Great Southwest Air Christmas Meltdown

On pages 608-9 of your text, the Global Company Profile for Ch. 15 features Alaska Airlines and is titled “Scheduling Flights When Weather is the Enemy.” The case study in the Linear Programming chapter (Module B) discusses how LP software drives staff and plane scheduling in that industry.

Southwest Airlines software is an aging in-house system called SkySolver, which as The Wall Street Journal (Dec 29, 2022) writes: “This Christmas, SkySolver not only didn’t solve much, it also helped create the worst industry meltdown in recent memory.” SkySolver was overwhelmed by the scale of the task of sorting out which pilots and flight attendants could work which flights. Crew schedulers instead had to comb through records by hand.

Even as SkySolver tried to solve one set of problems, new ones would emerge. Crews and planes were out of place. Phone lines jammed up, and Southwest staffers trying to get assignments couldn’t get through to the scheduling department. The airline was scrambling just to figure out where its crew members were located.

Southwest canceled more than 13,000 flights between Dec. 22 -27, while stranding passengers and bags across the country. This isn’t the first time that a disruption has ballooned at Southwest, and the carrier’s struggle shows how its increasingly complicated network needs a better technology foundation.

Pilots for years complained that SkySolver often spits out fixes that don’t make much sense, sending crews on circuitous journeys around the country as passengers to meet flights, a practice known as “deadheading.”

By Dec. 26, Southwest realized it needed a full reboot. In an effort to get pilots, flight attendants and planes into position, the airline took draconian measures. It canceled two-thirds of its planned flights for multiple days, and locked up seat inventory on its website so customers couldn’t buy tickets for a flight that might ultimately be canceled.

Unlike many rival airlines, Southwest’s planes generally hop from one city to another, rather than orbiting a major hub. That approach lets Southwest maximize use of its planes and crew, but the daisy chain structure also makes its network more delicate—problems in one corner of the country can be difficult to contain.

Classroom discussion questions:

Why did the scheduling software fail?
What options do operations managers have at this point?

OM in the News: Amazon’s New Delivery Route Algorithm

Amazon expects its new delivery route algorithm will help it avoid millions of miles driven this year after deploying it across the entire U.S., according to Supply Chain Dive (Aug. 31, 2022). The Customer Order and Network Density Optimizer (Condor) algorithm assesses customer orders before they leave a fulfillment center to identify the most effective shipping options. It began running in a few Amazon delivery stations in January before going live nationwide. “We can enable carriers to deliver more packages to more customers on time, while reducing miles driven and carbon emissions from fuel,” said an Amazon exec.

Amazon vans delivering packages in Hicksville, N.Y.

For Amazon, a decision on how a customer order should be fulfilled is made quickly based on item location and other factors. However, there can be a 5- to 6-hour gap from when an order is placed to when the fulfillment center begins processing the order, and the plan may be updated if a nearby customer places an order later in the day.

This period prior to fulfillment gives Condor a chance to reevaluate the original decision multiple times to see if a more optimal route can be created.

Condor takes into account the entire geographic area a delivery station serves while determining how orders are split into shipments and the sourcing fulfillment center for each shipment. The program’s “breakthrough” is that it is able to reduce the number of possible routing decisions to less than 10 per geographical block, despite the complexity involved.

Delivering the same number of packages in fewer routes and miles driven helps parcel carriers save on last-mile shipping costs, a big reason why UPS and FedEx have launched their own initiatives in this regard. UPS’ pilot involves holding an order for as long as the service agreement allows until it can match another shipment to the same address. FedEx is consolidating routes run by its separate Express and Ground companies via its “Network 2.0” plan.

Classroom discussion questions:

Explain how Condor works.
What is meant by “last-mile” shipping costs?

Guest Post: Southwest’s Airplane Disruptions

Today’s guest Post comes from Prof. Howard Weiss, who upon returning from Europe, observed an aggregate planning problem.

From October 10 through October 12, 2021 Southwest canceled more than 2400 of its flights due to bad weather and air traffic control problems in Florida, reported The Wall Street Journal (Oct 21, 2021). Your Heizer/Render/Munson Aggregate Planning chapter (Ch. 13) discusses planning for the airline industry and more specifically about airlines that use hubs. But Southwest does not use hubs. Rather it operates a very complex spider web-like point-to-point route network, and thus a flight between two cities in California can be impacted by bad weather in Florida, for example. According to Southwest’s COO, “40% to 50% of the airlines’ planes flow through Florida nearly every day, and many crews change there. So when we have a disruption, a significant disruption, in Florida, it tends to spread to our entire network.” By Friday night, he said, the airline had “well over” 100 planes and crews “that weren’t where they were supposed to be.”

Your textbook suggests that for successful Aggregate Planning there should be:

Accurate scheduling of labor hours.
An on-call labor resource
Flexibility of individual worker skills.
Flexibility in output rate

These suggestions are difficult for any airline to follow but even more difficult for Southwest. Of course, the cancellations led to a large increase in the number of customers trying to reach Southwest and because telephone staffing did not increase this led to waits of several hours to reach customer service.

Classroom Discussion Questions

What could Southwest do to appease customers whose flights were cancelled?
How does Southwest’s on-time performance and cancellation rate compare to other airlines’ performance in general?

OM in the News: Why Do the Empty Planes Keep Flying?

An Atlanta-to- Baltimore Delta flight on April 20

Many of us are wondering why airlines continue to fly nearly empty airplanes. It seems like they are burning cash, fuel and goodwill. Don’t they know what they’re doing? They do, says The Wall Street Journal (May 7, 2020).

Airlines scrambled in March and April to ground as many trips as they could. But they still found themselves locked into flying many trips with hardly any passengers for a number of OM reasons that show just how complex airline schedules are and how hard the choices are that OM executives must make.

Airlines have grounded 3/4 of their capacity, and it still hasn’t been enough. More than 90% of traffic has disappeared. For American, 99% of flights have been less than 20% full. Besides critical travel for funerals, medical reasons, etc., there are often operational reasons that they need to fly with 2-3 people on board.

A flight might have to go with 2 fliers because a later flight in that plane’s schedule had 60 waiting. Sometimes it isn’t the aircraft that had to get there, but the crew needed somewhere else for later flights. Some aircraft need to get to a maintenance base for overnight routine work. Overnight parking is in short supply at many airports housing grounded planes. So a jet may need to make the last trip of the day with 1-2 passengers just to get to its assigned parking space. And a few passengers might luck out because the plane has a large cargo payload–often medical supplies.

Now a growing concern is that there will be heavier loads on flights where passengers won’t be able to socially distance. Adding to the complexity is a requirement in the federal airline bailout (Cares Act) that requires airlines to maintain service to all the cities they currently serve. Finally, because of the complexity of schedules, airlines need to figure out which aircraft to put in storage while facing uncertainty over travel patterns in the pandemic.

Classroom discussion questions:

List some of the operations issues airlines are now facing.
What percent of your class is willing to fly in the next 4 weeks?

OM in the News: The NBA’s “Dirty Little Secret”

“There’s not a factory on the planet,” says one scientist, “that would move shift workers the way we move NBA players.”

IT’S THE AFTERNOON of Feb. 26, during a 3-games-in-4-nights stretch, and Miami Heat center Hassan Whiteside is on a roll. Tomorrow night, his Heat will host the Golden State Warriors, then fly to Houston to face the Rockets on Feb. 28. But now he’s rattling off what time the Warriors game will end (10 p.m.), when they’ll board their flight (after 11:30), when they’ll land in Houston (2 a.m.) and arrive at the hotel (3 a.m.) before playing the Rockets later that day.

Sleep matters, Whiteside says — it matters a lot. It “could be the difference between you having a career game or playing terrible.” Is it possible within the current NBA schedule to obtain consistent, quality sleep? “Nah,” Whiteside says. “It’s impossible. It’s impossible.”

Fatigue has long been a reality of life in the NBA, a league with teams that play 82 games in under 6 months and fly up to 50,000 miles per season — enough to circle the globe twice, reports ESPN.com (Oct. 14, 2019). Over the 2018-19 season, the average NBA team played every 2.07 days, had 13.3 back-to-back sets and flew the equivalent of 250 miles a day for 25 straight weeks.

Despite the league’s best efforts — lengthening its schedule in recent years, reducing back-to-backs for 5 straight seasons (down to an average of 12.4 per team in the coming season), eliminating 4-in-5 stretches, reducing the nationally televised games that tip off at 10:30 p.m., creating more rest days — sleep deprivation remains “our biggest issue without a solution. It’s the dirty little secret that everybody knows about,” says an NBA exec.

“I think in a couple years,” Tobias Harris says, “sleep deprivation will be an issue that’s talked about, like the NFL with concussions.” During the season, its estimated that players get 5 hours sleep per night. Chronic sleep loss has been associated with higher risk for cancer, diabetes, obesity, heart disease, heart attacks, Alzheimer’s, dementia, depression, stroke, psychosis and suicide.

Classroom discussion questions:

How is this an OM issue?
What can be done to alleviate the problem, and what dangers are possible if it isn’t addressed?

OM in the News: Using OM to Fix School Bus Routes

Last year, more than 30,000 students in the Boston Public Schools rode 650 buses to 230 schools at a cost of $120 million. In hopes of spending less this year, the school system offered $15,000 in prize money in a contest that challenged competitors to reduce the number of buses. The winner, reports The Wall Street Journal (Aug. 12-13, 2017), was MIT’s Operations Research Center, which devised an algorithm that drops as many as 75 bus routes.

The school system says the plan, which will eliminate some bus-driver jobs, could save up to $5 million, 20,000 pounds of carbon emissions a day and 1 million bus miles each year. The computerized algorithm runs in about 30 minutes and replaces a manual system that in the past has taken transportation staff several weeks to complete. “They have been doing it manually many years,” said the MIT Center co-director. “Our whole running time is in minutes. If things change, we can re-optimize.”

The task of plotting school-bus routes resembles the classic exercise known as the Traveling Salesman Problem (TSP), where the goal is to find the shortest path through a series of cities, visiting each only once, before returning home. Although we don’t cover TSP in the written text, we do so in our MyOMLab Tutorial called Vehicle Scheduling & Routing.

Classroom discussion questions:

Provide other applications of TSP.
How does this scheduling problem differ from that of airlines scheduling planes and crew?

OM in the News: Scheduling Star Wars to Maximize Profits

A Darth Vader serves popcorn to fans on opening night of Walt Disney Pictures And Lucasfilm's "Star Wars: The Force Awakens" — A Darth Vader serves popcorn to fans on opening night of Walt Disney’s “Star Wars: The Force Awakens”

Star Wars: The Force Awakens is playing in more than 4,100 U.S. theaters this week. And with many cineplexes running non-stop shows in multiple theaters, that adds up to tens of thousands of showings. For Disney, the math is simple: Sit back and watch hundreds of millions roll in. But for movie theaters, reports Marketplace Business (Dec.18, 2015), scheduling Star Wars in order to maximize profits requires some pretty elaborate calculations. There are many factors that come into play when setting a schedule: estimated ticket sales, number of screens devoted to the movie, time it takes cleaning crews to get in and out, and weather and holidays.

While it might seem easier to schedule Star Wars every 15 minutes, this actually is more difficult. First, because the movie is being released in 2D and 3D, theater owners have to predict the demand for both. They also have to decide which other movies they will stop showing to make room for Star Wars. (Theaters don’t start to cash in big on ticket sales until a movie has been playing longer.)

The real money is made at the concession stand. Theaters keep about 5% of ticket sales. Profit margins on concessions can run as high as 85%. It’s essential theaters schedule their movies to start and end at different times. They don’t want too many people coming and going simultaneously which create crowds that slow down the concession counter. Get it wrong, and they will have crowds waiting to buy tickets, in line for the bathroom–and not buying Junior Mints. Some theaters use algorithms and software to do their scheduling. Others still take their cue from Han Solo and go with their gut.

Classroom discussion questions:

What scheduling tools could theaters use?
Why is this such a complex issue?

OM in the News: At UPS, the Algorithm is the Truck Driver

“Here’s a math problem for you,” writes The Wall Street Journal (Feb.17, 2015). Each United Parcel Service driver makes an average of 120 stops per day. There are 6,689 times 10 to the 195th power alternatives for ordering those stops! Which option is the most efficient, after considering variables such as special delivery times, road regulations, and the existence of roads that don’t appear on a map?

Even if an optimal answer exists, the human mind will never figure it out. And while experts at UPS have been giving the problem their best shot for more than a century, the company is shifting that work over to a computer platform, with 1,000 pages of coding, called Orion, which is 10 years in the making. Considered the largest operations research project in history, the $200-300 million algorithm was written by a team of 50 UPS engineers.

Orion consists of many components, including a “traveling salesman” algorithm, a tool that calculates the most efficient path between a variety of points, and geographic mapping. None of the solutions that Orion spews out are big or dramatic. It is all about saving $1-2 here and there. But in a network with 55,000 routes in the U.S. alone, that adds up. E-commerce has shifted more and more of UPS’s delivery stops to residences, and those packages are expected to make up 1/2 of all deliveries. It is a radical routing change from 15 years ago, when drivers would drop off several packages at a retailer.

Orion is expected to save the company $300-$400 million a year once it is fully implemented in 2017. (UPS saves $50 million a year by reducing by 1 mile the average daily travel of its drivers.) But reaction to Orion is mixed. For example, some drivers don’t understand why it makes sense to deliver a package in one neighborhood in the morning, and come back to the same area later in the day for another delivery. But Orion often can see a payoff, measured in small amounts of time and money that the average person might not see.

Classroom discussion questions:
1. Why is Orion so important to UPS?

2. Why is the software so complex?

OM in the News: Scheduling Zappos’ Call Center Employees

Last September, Zappos’ CEO Tony Hsieh was wandering the halls of the online retailer’s Las Vegas headquarters and noticed that the customer service center’s walls were covered—floor to ceiling—with sheets of printer paper. He had stumbled across the scheduling method for the center’s 540 employees, who respond to the 10,000 customer inquiries the online retailer receives every day. Employees choose their shifts in order of seniority, by writing their names on sheets of paper listing the shifts they want. “It was like how I signed up for college courses before I could do it on a computer,” says a senior manager at Zappos.

“The old-school, paper-and-pencil process didn’t sit well with Hsieh,” writes Fortune (Jan. 28, 2015), “who is known for his devotion to customer service.” (The Amazon-owned company aims to answer 80% of customer inquiries within 20 seconds.) The wasteful manual sign-up process is now being replaced with Zappos’ Open Market, a newly created online scheduling platform that allows workers to set discretionary hours and compensates them based on an Uber-esque surge-pricing payment model: hourly shifts with greater caller demand pay higher wages. The goal of Open Market is to create a “free-market system,” and strike a balance between the rigidness of customer service center scheduling and what the company says is its dedication to giving employees time to pursue other opportunities at Zappos. Everyone receives at least 10% flexible time, so during a 40-hour week, employees would have 4 hours to play with. They could choose to not work during those hours or they could fulfill them whenever they liked by tacking them onto the start or end of a workday or by coming into the office on a scheduled day off.

Employees decide when to work with the help of Open Market’s real-time metrics algorithm that shows customer demand, as measured by the wait time of the longest-holding customer, and the accompanying compensation rates. The longer the hold time, the higher the customer demand, the more the employees working that shift would get paid. The idea is to tie compensation for the employees—who earn an average of $14.50 per hour—into the Open Market model and pay them a range of hourly rates based on demand.

Classroom discussion questions:

1. What are the advantages of the new scheduling system?

2. How does Open Market differ from employee scheduling systems you are familiar with?

OM in the News: Airlines and the Winter Storms

Workers at Boston's Logan Airport faced a mountain of snow last week — Workers at Boston’s Logan Airport faced a mountain of snow last week

As we discuss in the Global Company Profile on Delta Airlines that opens Chapter 15, airlines have made rapid improvement in storm recovery. Using new tricks, techniques and conservative strategies to position airplanes and employees before storms hit, flights can resume quickly as soon as runways clear. The change is a reaction to stiffer regulations, including stronger rules on pilot rest, and customer outrage on tarmac delays.

New software in use at several big airlines reschedules crews in minutes, sparing dispatchers hours of manual puzzle-solving, reports The Wall Street Journal (Feb.5, 2015). Last week, United got planes back into its Newark, N.J., hub and other Northeast airports on Tuesday night and was back at full strength by Wednesday afternoon, using only 11 reserve pilots. In past storms, that number would have been in the hundreds. The airline keeps 15-18% of its 10,000 pilots on reserve duty to fill in when pilots get sick or delays throw off schedules. United, like other airlines, sometimes ran out of reserve pilots and flights would have to be canceled days later in sunny weather.

United has long used software to reschedule planes, then given dispatchers the task of manually finding available crews. But this time, a new computer program rescheduled pilots sidelined by snow. “That optimization is so beautiful,” says United’s VP. “Last year, we had multiple events where the recovery dragged on for days.”

Airlines say tarmac-delay rules, where the government imposes big penalties on carriers for leaving passengers marooned on grounded airplanes, have forced them to avoid the risk of flying close to the storm. Shutting down early turns out to be better for passengers and airlines. Fewer passengers get stranded at airports because they never start trips or know ahead of time they won’t get home so they avoid the airport. United actually ferried 7 planes empty into Newark Tuesday afternoon so they’d be ready for early-morning departures Wednesday. By 4 p.m. Wednesday Boston was back to normal.

Classroom discussion questions:

1. Why is the airline operations center so critical during weather disruptions?

2. What does optimization software do for the airlines?

OM in the News: American Airlines Returns to “Peak” Scheduling

Shorter connecting times mean runs of up to 1.1 miles in Miami's airport — Shorter connecting times mean runs of up to 1.1 miles in Miami’s airport

“American Airlines is making its Miami hub more hectic—on purpose,” writes The Wall Street Journal (Sept. 11, 2014). Instead of spacing flights evenly throughout the day, the airline just started bunching them together. The change restores an old format of “peak” scheduling, grouping flights into busy flying times followed by lulls when gates are nearly empty. American next year will “re-peak” schedules at its largest hubs in Chicago and Dallas-Fort Worth.

Airlines shunned peak schedules at hubs more than a decade ago because they meant higher costs such as more people and equipment, created too many delays and forced passengers to sprint through terminals to make connecting flights. Recently though, the industry has gravitated back to peaks and valleys as a way to fill seats and generate more revenue. “An additional person per flight will make a difference,” said American’s CEO. The company will gain $200 million more a year from re-peaking its schedules at hubs.

But travelers may have even less time to make flight connections or to eat. And airlines, airports and federal agencies are re-evaluating how they manage baggage, cleaning crews and security checkpoints with the new highs and lows in foot traffic. Peak scheduling packs planes better because it creates more possible itineraries, with shorter connection times. In Miami, 42 flights depart between 9 and 10 a.m. Then between 10 and 11 a.m., only a handful are scheduled to take off. The process repeats during the day with 10 “banks” of flights that fill about 45 gates at a time.

There are added costs to re-peaking. American hired 67 more gate agents and 150 baggage handlers and other ground workers. It had to purchase more belt-loaders, dollies and tugs that push planes out from gates. There are other pitfalls to airlines’ clumped schedules. If bad weather hits at the wrong time, diverted flights and missed connections can cause widespread delays.

Classroom discussion questions:

1. What are the advantages and disadvantages of this approach?

2. What other OM ideas could American use to increase efficiency?

OM in the News: Delta’s Unorthodox Scheduling System

Delta's Control Room — Delta’s Control Room

“The crew of Delta Air Lines Flight 55 last Thursday couldn’t legally fly from Lagos, Nigeria, to Atlanta unless they waited a day due to new limits on how much pilots can fly in a rolling 28-day period,” writes The Wall Street Journal (April 3, 2014). The trip would have to be canceled. Instead, Delta headquarters told the captain to fly to San Juan, which they could reach within their duty limits. There, two new pilots would be waiting to take the Boeing 767 on to Atlanta. The plane arrived in San Juan at 2:44 a.m., quickly took on fuel and pilots, and landed in Atlanta only 40 minutes late.

The episode, unorthodox in the airline industry, illustrates the fanaticism Delta now has for avoiding cancellations. Last year, Delta canceled just 0.3% of its flights. That was twice as good as the next-best airlines, Southwest and Alaska, and five times better than the industry average of 1.7%.

As it cut cancellations with a more-reliable operation, overall on-time arrivals improved and Delta has fewer delays. Managers in Delta operations center (featured in our Global Company Profile in Chapter 15) move planes, crews and parts around hourly trying to avoid canceling flights. How well an airline maintains its fleet and how smartly it stashes spare parts and planes at airports affect whether a flight goes or not. Delta’s new analytical software and instruments that can help monitor the health of airplanes and predict which parts will soon fail. Empty planes are ferried to replace crippled jets rather than waiting for overnight repairs. Typically the airline has about 20 spare airplanes of different sizes each day. About half are stationed in Atlanta and the rest spread around other domestic hubs and two in Tokyo.

Classroom discussion questions:

1. Why have Delta’s operations managers focused on cancelling fewer flights?

2. How does Delta’s fleet age (one of the oldest in the industry) impact this strategy?