United Parcel Service is a global organization, with 424,000 employees and nearly 100,000 vehicles. UPS drivers typically make between 120 and 175 drops per day. Between any two drops, drivers can take a number of possible paths. With 55,000 routes in the Unites States alone, the total number of possible routes is inconceivably vast. Clearly, it is in the best interest of UPS and its drivers to find the most efficient routes. Therefore, any tiny amount of efficiency that can be gained in daily operations yields significant improvements to the company's bottom line. Essentially, little things matter a great deal to UPS.

In addition, UPS must manage a low-margin business as well as a unionized workforce that is compensated at the high end of the industry scale. Significantly, rival FedEx uses independent contractors for its ground network. Consequently, FedEx does not have the burden of expensive employee benefits packages.

Another problem for UPS is that the increase in electronic commerce has shifted an increasing number of UPS's delivery stops from retailers to residences. In fact, UPS expects that residential deliveries will make up half of the company's total deliveries by 2018. Historically, drivers would drop off multiple packages at a retailer. Today, they must make scattered stops to drop off packages at individual houses in a neighborhood. This process involves more routes and is more time consuming.

The Solution

For decades UPS has been using three types of analytics to produce efficiencies:

Descriptive analytics asks, Where are we today? Predictive analytics asks, With our current trajectory, where will we be tomorrow? Prescriptive analytics asks, Where should we be tomorrow? As UPS has moved from descriptive to predictive to prescriptive analytics, its data needs have increased, the skill set of its people has increased, and the business impact of analytics has increased. We consider these developments below.

Descriptive Analytics. UPS implemented descriptive analytics in the 1990s when the company provided its drivers with hand-held computers, called Delivery Information Acquisition Devices (DIADs). The DIADs enabled UPS to capture detailed data that measured the company's current status. For example, the company measured driving variables in hundredths of a second. Their reasoning was that if they could reduce one mile per driver per day in the United States alone, that process would add up to $50 million to the bottom line annually.

One drawback to the DIAD system was that the data were scattered across various locations. Specifically, some of the data were with employees, some were located in corporate repositories, some were contained in Excel spreadsheets, and some were distributed throughout the company. However, UPS did not have a predictive model that could help the company tomorrow.

Predictive Analytics. To address this problem, in 2003 UPS deployed predictive analytics with its Package Flow Technologies system. With this system, drivers started the day with a DIAD that detailed the packages they were to deliver and the order in which they were to deliver those packages. The DIAD became the drivers' assistant. The system enabled UPS to reduce total delivery driving by 85 million miles per year. That process saved the firm 8.5 million gallons of fuel, and it saved the planet from 85,000 metric tons of carbon dioxide going into the atmosphere.

However, drivers had to provide different services from the same vehiclefor example, deferred service and premium service. They had some packages that had to be delivered by 10:30 AM, some that had to be delivered by noon, and some that had to be delivered by 2:00 PM. Drivers therefore had to decide how they were going to service those customers. With so many variables to consider, it was practically impossible for drivers to optimize their routes.

Prescriptive Analytics. UPS realized that it needed to take analytics to the next level. So, in mid-2012, the company began deploying its On-Road Integrated Optimization and Navigation (ORION) system. ORION reorganizes the drivers' routes based on today's customers, today's needs, and today's packages, and it designs deliveries in a very specific, optimized order. ORION takes into account UPS business rules, maps, what time drivers need to be at specific locations, and customer preferences.

When UPS drivers are on the road, they usually travel at speeds of 2025 miles per hour. Therefore, every mile reduced equates to a savings of 23 minutes. Because ORION shortens routes by 78 miles per day, this savings enables UPS to deliver more packages.

ORION enhances UPS customer service with more efficient routing, and it allows UPS to offer innovative services and customized solutions. An example of this type of service is UPS My Choice, which gives customers a one-day alert for the time a package is arriving, and allows them to control the timing and location of the delivery. By mid-2015, UPS My Choice had almost 13 million members.

At the beginning of a shift, a UPS driver checks his DIAD, which displays two possible delivery routes. One route uses ORION, and the other uses UPS's traditional combination of work rules, procedures, and analytic tools. Drivers can choose either route, but if they decide not to use ORION, then they will be asked to explain their decisions. Driver reaction to ORION has been mixed. The experience can frustrate some drivers who might not want to surrender their autonomy or who might not follow ORION's logic in designing their routes.

The Results

In 2013, ORION saved 1.5 million gallons of fuel (on top of the 8.5 million gallons the firm was already saving). UPS gained those savings despite the fact that only 20 percent of its 50,000 drivers had adopted the system. UPS plans to complete the deployment of ORION by the end of 2016. When ORION is fully implemented, the system is expected to reduce the distance driven by its drivers by 100 million miles annually. Further, ORION will produce annual savings of $300 million as well as an additional 10 million gallons of fuel. In addition, ORION will produce further environmental benefits and cost reductions when UPS vehicles outside the United States are equipped with the technology.

UPS continues to look into the future. Interestingly, ORION provides a natural link to driverless vehicles.

However, ORION does not yet perform certain functions. For example, when drivers leave in the morning, the route they have in their DIAD does not change, meaning the system does not update routes if something goes wrong. In addition, ORION also does not take traffic or weather into account. UPS plans to integrate these features into subsequent upgrades of ORION.

1.Explain how DIADs were a descriptive analytics solution for UPS. 2.Explain how the Package Flow Technologies system was a predictive analytics solution for UPS. 3.Explain how the ORION system was a prescriptive analytics solution for UPS. 4.Describe another potential application for the UPS ORION system. That is, what is the next question that UPS managers might ask of the ORION system?