Big Data Delivers Big Results at UPS


Randy Stashick, president of Engineering at UPS, spoke at the Production and Operations Management Society Conference on the importance of big data and operations research when designing and implementing breakthrough changes.

Since we’re going to be talking about Big Data today, let’s start with a big number.

Any idea what that number represents? Let me tell you a couple of things it does not represent.

It’s not the number of hotdogs the famous Varsity restaurant, just up the street from us, has sold since opening in 1928. Nor is it the number of cars on Atlanta’s infamous interstates at 5 o’clock on a Friday afternoon.

Actually, that number, 199 digits in all, represents the number of discrete routes a UPS driver could conceivably take while making an average of 120 daily stops.  

Now, if you really want to get crazy, take that number and multiply it by 55,000. That’s the number of U.S. routes our drivers are covering each business day.

To display that number, we’d probably need that high-definition screen at AT&T Stadium in Dallas, where the Cowboys play.

But somehow UPS drivers find their way to more than nine million customers every day, to deliver nearly 17 million packages filled with everything from a new iPad for a high school graduate in Des Moines, to insulin for a diabetic in Denver, to two giant pandas relocating from Beijing to the Atlanta Zoo.

How do they do it? The answer is operations research. Big Data is delivering big results at UPS. It’s an amazing story and one I’d like to share with you this morning.

As impressive as that number on the screen, there’s another number that’s even more impressive to those of us at UPS.

It’s the number 1. As in, the one route that is the fastest, most precise and most cost-effective route our drivers could take to service all of their customers.

That’s the one we’re focused on and we’ve spent years, millions of dollars and countless hours trying to find it. We think we have. More on that breakthrough solution in a minute.

First, some context, along with a little history.

My mother used to tell me the devil was in the details. That was a good lesson because I work for a company that lives in the details. And those details can really add up.

For example, we have more than 96,000 vehicles on the roads most days servicing millions of customers. We’re very much aware that time, and miles, mean money.

We know that a reduction of just one mile each day per driver over the course of a year saves the company up to $50 million dollars.

By the same token, when a single driver shaves a single minute of idling time across our entire fleet, we save about $515,000 dollars in fuel alone.

That same minute is worth about $14.6 million dollars in operational savings. Those are details that go straight to the bottom line.

With so much riding on the details, you can see why we love operations research and analytics.

Before we get too deep into our 21st century story of operations research and package flow technology, I want to take you back a few years. Back to the start of our company.

Because to appreciate our application of Big Data, you really need to understand the journey that’s brought us to this point.

UPS got its start in 1907 in Seattle, with a few young boys, a few bicycles and a visionary. 

The visionary was our founder, Jim Casey, himself a teenager, only 19 when he started the company in the basement of a Seattle storefront.

But he had big dreams. And he wasn’t afraid of change. He proved that by reinventing the company time and again, long before anyone ever used the term “reinventing the company.”

The first reinvention followed widespread adoption of the telephone. That could have put our company, which was based on delivering messages, out of business.

But Jim went to the big department stores in Seattle and convinced them to let his company deliver packages to their customers.
Remember, this was before most people had cars. So most big stores had their own fleet of delivery vehicles.

Jim convinced the retailers that we could do it faster and save them money in the process. That was really the beginning of the UPS business we know today.

There have been several other reinventions over the years, including the decision in the late ‘80s to start our own airline. UPS now operates some 500 aircraft, making us one of the world’s largest.

The most recent reinvention came in the mid '90's, as the Internet was morphing from the, “Information Superhighway”, remember that quaint term? to the digital über sales channel we know today.  

Driven by the mega-forces of globalization and technology, UPS built the largest and most sophisticated technology infrastructure in the transportation industry.

Today, in Mahwah, New Jersey, and in Alpharetta, Georgia, about 10 miles north of here, 10 IBM mainframes operate around the clock.

They’re capable of processing 27 million instructions every second and tracking 16 million packages every day.

They collect and distribute information from drivers’ hand-held computers from around the world. Then they use that information to coordinate the operations of our global fleet of vehicles, as well as an entire airline.

Globalization and technology have combined to dramatically alter our business. For decades we moved goods, information and funds along independent pathways because that was our only choice.

But in a smaller, flatter world, and with the help of technology, we’ve integrated all three of those flows so they move together, in sync. We call it synchronized commerce.

Synchronized commerce has taken us from providing a transportation service, to providing a business solution. We’ve evolved from a trucking company with technology, to a technology company with trucks.

This latest reinvention of our company was possible only because of operations research.

Operations research is embedded in our corporate DNA. As far back as the ‘50s, George Smith, our company’s second CEO, was praising its benefits.  

Here’s what he said about it:
“The ability to make decisions is the power to manage. Without operations research, we could only analyze our problems intuitively.”

Today, we proudly underwrite the UPS George D. Smith Prize in his memory to strengthen ties between industry and higher education.

This year’s winner is an organization known as Leaders for Global Operations at MIT. The prize was announced at the Franz Edelman Gala during the INFORMS Analytics Conference in late March.

This organization is the nation's leading graduate dual-degree program in engineering and management innovation.

What George Smith understood so well decades ago is that data collection and analysis provide the connective tissue for efficient, reliable operations.

That truth has become even more obvious as the company evolved over the last half century. We’re convinced that breakthrough change is impossible without operations research.

Over the decades, our data structure has matured to model our operations.

One example is our shared transportation network for standard ground and air packages.

Unlike our competitors, the same UPS driver delivers air and ground packages out of the same delivery vehicle, no matter which level of service the customer has chosen.

This adds complexity to our model, but efficiency to our operation.  It also reinforces the strong service relationship between our customers and their drivers.

Our data infrastructure is designed to know the location of every package, where it’s going, when it needs to be there and whether it got there on time.

Needless to say, when you’re delivering nearly 17 million packages a day, the data is complex and extensive. But, as I said, Big Brown thrives on Big Data.

We utilize this rich data infrastructure to drive innovation throughout our operations. We take a team approach, evaluating input from everyone involved – particularly our drivers, who provide invaluable insights from their unique perspective.

In general, our process for operations can be summed up in four words: Plan … Do … Check … Act.

  • We collect data, analyze the operational situation and forecast future demand patterns.
  • From the forecast, we develop a plan for optimum efficiency.
  • We execute the plan, all the while monitoring and adjusting in real time.
  • We close the loop with analysis and begin the process again.

We touch all the bases, and we do it day after day. So let’s look at how this process plays out on a daily basis. Remember, time and miles are precious commodities in our world.

We make every effort to load our delivery vehicles, we call them package cars,to match the order in which the packages need to be delivered. Business deliveries are prioritized over residential.

Not long ago, loading packages in the most efficient manner possible was a weakness in our operation.

It required a preloader, the person responsible for loading packages, to memorize a delivery route, along with all of the addresses assigned to his or her route. Obviously, a complex job.

Pre-loaders went through extensive training. But that didn’t keep our operation from being affected by something as simple as an unexpected sick day.

Considering all of the challenges, our preloaders did an amazing job. But we knew this basic process was due for an overhaul.

The makeover started with what we call the Pre-Load Assist Label, or PAL, which is printed just before the preloader touches the package for the first time.

This label simplifies the preloader’s job by pinpointing exactly where the package needs to go on the vehicle.

Thanks to PAL, anyone in this room could become an effective preloader in a matter of minutes. By the way, we’re hiring if anyone is interested. 

PAL is just one component of the package flow process, which begins with data acquisition.

Another tool drivers use is something called the Delivery Information Assistant Device or DIAD.

Over the last 20 years, DIADs have evolved from a rudimentary electronic replacement for a basic clipboard to a sophisticated technology assistant for our drivers.  

The DIAD plays a critical role, supplying data to drivers so they can make better-informed decisions, including the most efficient delivery order.

The DIAD also knows the commitments we make to customers and communicates those promises to the driver.

We believe our drivers are the best in the industry. Millions of customers who appreciate their reliability and personal service back up that claim.

But they’re even better, more efficient and able to handle increased package volumes and services, because of the DIAD.

The DIAD has proven the truth of the formula, E plus I is less than E, which was derived from an award-winning paper UPS co-authored with the University of Georgia. In its simplest form, if you take Energy and add Information, less Energy is required for the same task.

PAL and DIAD are supported by a sophisticated delivery forecasting and planning engine that predicts future demand patterns and creates daily delivery plans.

Our systems analyze these plans and identify inefficiencies. It’s part of a dynamic process that can be adjusted up to the moment a driver leaves the sorting facility. 

Of course, we continue to monitor and adjust delivery plans and gather additional data long after the driver leaves the building.

Thanks to telematics, which work in conjunction with the DIAD, we’re constantly gathering data. We know where the driver is, what work has been completed, and what work is remaining.

More than 200 sensors in the vehicle tell us if the driver is wearing a seat belt, how fast the vehicle is traveling, when the brakes are applied, if the bulkhead door is open, if the package car is going forward or backing up, the name of the street it’s traveling on, even how much time the vehicle has spent idling versus its time in motion.

Unfortunately, we don’t know if the dog sitting innocently by the front door is going to bite.

We also gather a ton of diagnostic data.

For example, dead batteries discovered on Monday morning used to delay hundreds of drivers. Now, telematics help us continually monitor battery conditions and take preemptive action, like changing out or charging batteries so vehicles stay on the road.

The benefits gained from systems like package flow technology and telematics are undeniable.

Big Data has contributed to a reduction of 85 million miles driven per year, 8.5 million gallons of fuel and 85,000 cubic tons of carbon emissions. 

We’ve also achieved nearly 100 percent seat belt use from our drivers and reduced our preloader training time by 95 percent.

So where do we go from here? How do we build on a foundation that has always prioritized efficiency and continue to get even more efficient, add more services and improve what is already an impressive safety record?   

One obvious answer is that we will continue to rely on operations research supported by Big Data. We’ll keep asking questions like the ones you see on this slide.

We know from our own experience, as well as from a Gartner study, that business impact and analytics go hand in hand.

We’re now approaching the highest level of analytics maturity with what we call ORION.

ORION is an acronym for On-Road Integrated Optimization & Navigation.

It’s a technology that identifies the best way to deliver all of the packages in the vehicle while simultaneously meeting the needs of customers and our business. It’s route optimization on steroids.

ORION consists of more than 250 million address data points.
The software combines customers’ shipping requirements with customized map data that gives drivers precise routing instructions that reduce miles and our carbon footprint.
We began deployment of ORION last fall. We’re already seeing very significant fuel savings and major reductions in carbon dioxide emissions.
As I mentioned, UPS drivers deliver to an average of 120 stops a day – often more during our peak season around the holidays.

Their options are essentially endless. More than the age of the earth in seconds, to throw out just one gee-whiz fact. Humans can’t think of all the possible ways to deliver a route.

That makes ORION, according to analytics and business process expert Thomas Davenport, arguably the world’s largest operations research project.

ORION comes up with better routes because it considers literally millions of options faster than a driver can start his vehicle. Remember that big number from earlier? That’s the power of ORION.

David Pogue, host of PBS’s popular science program, NOVA, discovered the truth of that statement recently. Let’s take a look as David challenges the Goliath known as ORION.  

ORION got the better of David this time around, beating his answer by 10 miles.

What you didn’t see in the video is some of the detail behind David’s performance. 

David missed the service window on three out of four premium packages. He also missed the pick-up window on a couple of pickups. He also arrived at six businesses after they closed. Oops.

His experience emphasizes that simultaneously accounting for cost and service is extremely complicated. It also reminds us that the devil is in the details.

We started testing ORION in 2011. What we call the “ORION Curve” shows us what we expected: As more drivers are deployed, the performance of their package centers significantly outpaces centers not yet on the ORION platform.

So we’re obviously anxious to extend ORION’s benefits to our total driver workforce, which we expect to do by the end of 2016.   

While cost savings are certainly important, one of the most exciting aspects of operations research is that it facilitates new services.

Delivery Intercept and UPS My Choice are two great examples.

Delivery Intercept enables a shipper to intercept a package that’s already been shipped and have it returned or rerouted.

UPS My Choice gives customers more flexibility and control over their home deliveries. You can reroute, reschedule or authorize the release of your shipment online to best fit your needs.

Using My Choice, we can now send you a voice, text or email message the day before a delivery is scheduled so you know a package is on its way and can plan accordingly.

As we continue to develop ORION, and further leverage Big Data, future innovations like these seem almost endless.

We’ll certainly see more and more personalization, more condensed delivery windows and possibilities we haven’t yet imagined.

Operations research and Big Data are the most conspicuous levers we have pulled to improve performance and meet our customers’ needs.

But all the data in the world is irrelevant without skilled people to analyze it and influence changes in behavior.

We know that change management is an essential element of the gains we can realize from analytics.

That’s why we prioritize clear and consistent communication. For example, we consistently review the financial and environmental benefits of reducing idle time.

We also spend close to $500 million on training across the entire enterprise.

Without the commitment to communication, education and training, behaviors and results usually don’t change.

Big Data has generated some big results at our company.

Operations research has taken us from knowledge to wisdom.

Is clairvoyance, the ability to predict the future, a possibility? When we can predict a problem and find a solution before the customer is affected, then operations research and analytics are going to look like Kreskin is running the show.

Your field of study, what you’re teaching the George Smiths of tomorrow, is critical to the continued success of companies like UPS.

There are some major challenges, probably led by traffic congestion, that need to be tackled.

But whatever the challenge, we know this: Operations research will be central to solutions in this and other areas. Businesses worldwide are counting on it.

It’s been an honor for me to be with you today to celebrate the 25th anniversary of this conference.

On behalf of UPS, I want to congratulate the Production and Operations Management Society for its excellent work over the years.



For more information about this speech, contact:

Justin Luther

UPS Public Relations