2015 Broadband Tech Summit - Todd Westberg UPS Presentation

The Road to OptimizationUtah Broadband Summit

Todd Westberg – UPS Information Technology Director

UPS Facts

2

Founded in Seattle, Washington – 1907, HQ in Atlanta GA

435,000 employees, 220+ countries and territories

99,2892 package cars, vans, tractors, motorcycles, etc.

237 owned airplanes / 302 chartered (9th largest) -1,955 segments

Drive 2.5+ billion miles / year (U.S. alone)

Daily flight segments – 936 U.S. / 755 Int’l

Deliver 18 million packages a day / 8.2 million daily deliveries

UPS / Supply Chain & UPS Freight, UPS Stores and UPS Capital Corp

(Service parts logistics; technical repair and configuration, supply chain design and planning, returns mgmt.)

3

Chief Information Officer Dave Barnes

Number of technology employees 4, 443 + (1,009 FS technicians)

Physical Servers 21,484

Laptops and Desktops 211,278

UPS Global Telecommunications Network

Network Sites 3,479

Operating facilities served 1,990

UPS IT Support

Technology Facts

4

Data Centers Mahwah, NJ / Atlanta, GA

Mainframes 10

Mainframe Capacity (Millions of instructions/second – MIPS) 72,979

Petabytes of mainframe and UNIX storage 16.1

UPS Data Center

Technology Facts

5

www.ups.com Average Daily Usage

Page Views 33.4 million

Peak Day Page Views 50.8 million

Online Tracking Average Daily Usage

Tracking requests 58.2 million / business day

Peak Day Tracking requests 100.9 million

UPS Online

Technology Facts

6

Chasing Big Data

7

Big Data can transform businesses.

Big Data expands customer intelligence. Imagine telling your customer why they are calling

before they even ask the question!

Big Data improves operational efficiencies. Let’s start by looking at a big number:

Keep that number in mind while we look at the progression in data usage that we have

seen at UPS.

Any idea what that number represents?

Why is BIG DATA a BIG Deal?

668950291344912705758811805409037258675274633313802

981029567135230163355724496298936687416527198498130

8157637893214090552534408589408121859898481114389650

005964960521256960000000000000000000000000000

UPS has moved up the data maturity ladder starting with

methods and measurement, and moving to optimization.

70%

30%

16%

3%

UPS Evolution of

BIG DATA

DIAD(1991)

Telematics(2001 / 2006)

Methods and measurement

(1940’s)

ORION(2012)

Package Flow Tech.(2003)

EDGE(2015)

70%

30%

16%

3%

UPS Evolution of

BIG DATA

DIAD(1991)



(1940’s)

ORION(2012)


EDGE(2015)

<#>

UPS has been utilizing Big Data

for over 20 years

Delivery Information

Acquisition Device

(DIAD)

11

DIAD I

1990

DIAD II

1993

DIAD III

1999

DIAD IV

2004

DIAD V

2011

UPS has been utilizing Big Data

for over 20 years

12

Enhanced DIAD

Download (EDD)

13

Changing the DIAD from an Acquisition device to a

driver Assistant for making better decisions

Number of

Packages

Shelf

Location

Additional Activity(COD, signature, etc.)

Commit

Time

Address

Information

14

Keeping our promise to the customer

Enabling more services

Street # Street NameAdult Signature Required

3924 PLANDOME RD

------------------------------------------

1 Col Amt Ship # ID# Remark

1 #000000 675 CAL

Edit Dup Ship#<6 PreRec Ovrride

PkgInfo NonDel Notes Deliver

Park at Dock OnlyMailroom

Check in with SecurityUse freight elevator

No Indirect/Reroute/LA

Edit Dup Ship#<6 PreRec Ovrride

70%

30%

16%

3%

UPS Evolution of

BIG DATA

DIAD(1991)



(1940’s)

ORION(2012)



EDGE(2015)

16

Telematics

Telematics provides additional

sources of data

17

Speed Seat belt off in travel

Delivery while idling Route overlap

0 500 1,000 1,500 2,000 2,500

664463

664452

664049

664034

663547

663502

663481

663314

662783

661578

140120

134425

134416

134404

134370

134356

August 28 - Sep 1

Engine Speed to Road Speed - MBENZ - Roswell

10

20

30

40

50

01 Oct 06 01 Nov 06 01 Dec 06 01 Jan 07date

avgofpercent engine load engload_ucl

Percent Engine Load: Vehicle 131370

Predictive Failure Analysis

Telematics closes the loop by providing

detailed analysis

18

Integrated Data

Allows for Diverse Usage

Visibility on primary and behavioral characteristics that affect fuel consumption

1. Automotive 2. Safety 4. On-Road Performance 5. Work Measurement

Vehicle Diagnostics Data,

Fault Codes, and Usage

Cycles can be used to

anticipate part failure and

predict vehicle

breakdowns.

Through GPS, DIAD, and

Sensor Devices (Seatbelt,

Bulkhead Door, and

Reverse), a driver’s driving

habits can be monitored

and areas of improvement

can be displayed visually

on maps.

Dispatch Planning process

may be enhanced and

simplified through “Work

Area” Concept. Dispatch

planning & address

validation processes may

be enhanced and

simplified through GPS

data.

Combining GPS and DIAD

data displayed on maps,

allows us to monitor and

analyze daily driver travel

paths and highlight excess

miles and dispatch

inefficiencies.

Using GPS, DIAD, and

Map data, has allowed us

to enhance, automate and

simplify several steps in

On-Road time study and

work measurement

processes.

2. Safety 3. Dispatch Planning 4. On-Road Performance 5. Work Measurement1. Automotive

Ap

plic

ati

on

A

rea

s

70%

30%

16%

3%

UPS Evolution of

BIG DATA

DIAD(1991)



(1940’s)

ORION(2012)



EDGE(2015)

10- 20510- 245 10- 145

10- 005

10- 14010- 20010- 320

10- 240

10- 32510- 36510- 455

10- 405

10- 45010- 600 10- 400

10- 360

1118

To reduce driver time and miles, packages are loaded in a

general order of delivery based on preload knowledge

Prior to PAS

Simplification through Package Level Detail (PLD)

UPS Smart Labels act as a “trip ticket” to carry

packages through the network and the final mile

The preloader still loads the package in the General order of

delivery, but the package is now smart and tells the preloader

where it belongs.

1118

Breakthrough Change

UPS Data Infrastructure

Customer

Address(USPS)

Geography(Map)

UPS Proprietary

Regulatory

GPS

Vehicle

Service

The UPS model is designed to know where every package is,

where it is going, where it needs to go, and why.

25

Delivery Forecasting and

Planning

Each package planned on the

right vehicle for the right day

27

New Products

based on Data Architecture

• Delivery Alerts• Delivery Planner • Reschedule Delivery • Hold for Will Call• Authorize Shipment Release • "Leave at" Instructions (Driver

Instructions) • Leave with Neighbor• Deliver to a UPS Store• Deliver to Another Address• Approx. delivery time• Confirmed Delivery Window (2-hour)

28

Smart Stop

Improving Service

Every delivery at the right location

29

Real-time Request Execution

Through the use of our technology, our devices are

connected to our infrastructure AND to the customer.

30

The Impact (so far…)

85 million miles

driven reduced

8 million gallons

of fuel reduced

85,000 metric tons of

emissions reduced

Improvements in Safety, Service, and Performance

95% reduction

in loader

training time

Highest service

levels in

history

First of their kind

product offerings

>99.9% seatbelt usage 8 Billion fewer manual

entries by drivers

70%

30%

16%

3%

UPS Evolution of

BIG DATA

DIAD(1991)



(1940’s)

ORION(2012)

ORION(2012)


EDGE(2015)

32

On-Road Integrated Optimization & Navigation (ORION)

Operations Technology

Foundation

UPS Data Model

UPS’ ORION Algorithm

Additional Planning / Analysis Tools

UPS Technology

Infrastructure

Customized Map Data

ORION

ORION builds upon UPS’ rich technology foundation

33

• Optimizes a driver’s route using:

– Advanced mathematical models

– Data from planning systems

– Customized map data

• Accounts for business rules, customer needs, and

service commitments

• Allows “what if” analysis and decision making

• ORION can enhance customer service and reduce

the numbers of miles driven by determining the most

efficient delivery route.

Previous Route

ORION Route

On-Road Integrated Optimization & Navigation (ORION)

34

250 Million Global Data Points are used to optimize

routes based on exact delivery locations.

ORION

35

Why is this a breakthrough?

Why is this so hard?

Approximate age of the Earth (in Seconds):145,065,600,000,000,000

Number of ways to deliver 120 stops:

6,689,502,913,449,135,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,

000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,

000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,

000,000,000,000,000,000

36

ORION “sorts the list” from a General order to a

Specific order for the day’s specific conditions

37

Demonstrating Complexity

The ORION Test

What is the most cost effective way to

serve these customers?

38

What is the most cost effective way to

serve these customers?

Drivers have to worry about commit times, customer needs, and business rules etc…

How would one deliver all these points (there are 147)

Which route has fewer miles?

One has more than 10 miles

40

The ORION

Test

on NOVA

Working smarter rather than harder

41

Business rules add additional complexity

ORION reduces cost while satisfying all customer

and business needs

<#>

70%

30%

16%

3%

UPS Evolution of

BIG DATA

DIAD(1991)



(1940’s)

ORION(2012)


EDGE(2015)

EDGE(2015)

Breakthrough Change

UPS Data Infrastructure

Customer

Address(USPS)

Geography(Map)

UPS Proprietary

Regulatory

GPS

Vehicle

Service

What’s new for 2015? Listen to the following video as UPS moves

into the age of Enhanced Dynamic Global Execution systems.

EdgeVideo.mp4

<#>46

47

Big Data enabled the start of fully automated facilities.

These facilities feature new technologies for material handling and a sophisticated sorting

system.

They use the latest technologies to speed the sorting of "smart" packages and to reduce

the physical workload for employees.

Packages are immediately run through a scanner where information about the package is

captured by high-speed computers.

UPS's internally developed software uses that information to electronically guide the

package through the facility, directing it to the proper truck for delivery. The software,

designed to work in any automated UPS facility, even indicates the precise spot on the

truck where the package should be placed to ensure efficient delivery.

Automation

Let’s look at technology from a package’s perspective!

48

Today the price of automated Material Handling Systems has

come down

The cost of labor has risen

Competition for labor is very high.

Automated Material Handling Systems are faster and smaller

The return on investment is much better.

Key Factors make the Case for Automation

49

High Speed Package Singulators

Converting bulk flow into single file flow

BULK Package Flow Singulated Flow

Singulator

50

6-Sided Camera Decode Tunnels

Decoding the bar and Maxicode labels

51

High Speed Label Applicators

Applying PAS labels at high speed

52

High Speed Parcel Sorters

Sorting Parcels to their Destinations

<#>