Real-time (Big) Data in Freight Transport - Meeting the global trends

!

REAL-TIME (BIG) DATA IN FREIGHT TRANSPORT - MEETING THE GLOBAL TRENDS

Per Olof Arnäs Chalmers University of Technology @Dr_PO [email protected] about.me/perolofarnas Slides: slideshare.net/poar

Image: www.simonstalenhag.se

Northern LEAD Logistics Research Centre

Founded by: Chalmers University of Technology

University of Gothenburg Logistics and Transport Society LTS

Tomorrow’s logistics. We are finding the answers.

Around 70 researchersResearch centre for sustainable logistics

solutionsFive core research

groups

Organises, facilitates, disseminates highly relevant

logistics researchCollaboration between

Chalmers and University of Gothenburg

Professors 10 Visiting Professors 6 Associate Professors 5 Post docs 6 Faculty 14 PhD students 32 Total 73

Five core research groups

Physical Distribution

Production Logistics

Industrial marketing & purchasing

Logistics & Transport

Optimization

Doing some Sisyphus work by Kalexanderson on Flickr (CC-BY,NC,SA)

5 GLOBAL TRENDS

Source: PWC (google: pwc megatrends 2014)

Crowd by James Cridland on Flickr (CC-BY)

Megatrend #1

Demographic and social

change


Four Storms And A Twister by JD Hancock on Flickr (CC-BY)

Megatrend #2

Shift in economic

powerSource: PWC (google: pwc megatrends 2014)

Boston Downtown at Night by Werner Kunz on Flickr (CC-BY,NC,SA)

Megatrend #3

Rapid urbanisation


¡Rayos! by José Eugenio Gómez Rodríguez on Flickr (CC-BY,NC,SA)

Megatrend #4

Climate change and

resource scarcity


Megatrend #5

Technological breakthroughs


Demographic and social

change

Shift in economic

power

Rapid urbanisation

Technological breakthroughsClimate

change and resource scarcity

5 GLOBAL TRENDS

Stage Coach Wheel by arbyreed on Flickr

Development of transportation technology has been

fairly linear

…for the last 5500 years

We are in the middle of a gigantic exponential development curve

beginning

A new global eco system where new types of, knowledge based,

industries compete with traditional ones

http://jaysimons.deviantart.com/art/Map-of-the-Internet-1-0-427143215

Startups don’t compete with airlines...

by purchasing a bunch of planeshiring a bunch of pilots

and locking up a bunch of terminals at airports.

Quote: bryce.vc/post/18404303850/the-problem-with-innovation Image: Connecting the community, my Twitter strategy, and American Airlines at DFW by Trey Ratcliff on Flickr (CC-BY,NC,SA)

Startups compete with airlines by inventing videoconferencing.

Startups don’t compete with airlines...

by purchasing a bunch of planeshiring a bunch of pilots

and locking up a bunch of terminals at airports.

Quote: bryce.vc/post/18404303850/the-problem-with-innovation Image: Connecting the community, my Twitter strategy, and American Airlines at DFW by Trey Ratcliff on Flickr (CC-BY,NC,SA)

RESOURCE UTILISATION LOW

Source: Kent Lumsden



Safety imbalance Variation in resource demand

Chain imbalance Caused by the chain

Technological imbalance E.g. mismatch in equipment

Operational imbalance Goods and resource flow not compatible

Structural imbalance Uneven transport demand



Safety imbalance Variation in resource demand

Chain imbalance Caused by the chain

Technological imbalance E.g. mismatch in equipment

Operational imbalance Goods and resource flow not compatible

Structural imbalance Uneven transport demand

Several of these imbalances can be

reduced by reducing

uncertainties

But the biggest problem in transportation is time.

There is not enough of it. Ever.

In S

ea

rch

Of

Lo

st T

ime

by

bo

ge

nfr

eu

nd

on

Flic

kr

Strategic Tactical Operational Predictive

Time horizons Freight industry

Most (preferably all) decisions in the

transportation industry are made here. At the latest.

Uninformed, ad-hoc, and

probably non optimal,

decisions

Science fiction

The transport industry does not like real-time decisions.

At all.

Batch-handling

Zip codes Zones

Time-tables

DSC_9073.jpg by James England on Flickr (CC-BY)

Image: Alain Delorme, alaindelorme.com

The current model is focused on economy of scale and standardization

The current paradigm

So…

What are we doing about all

this?

Gartners Hype Cycle for Emerging Technologies

Augmenting humans with technology

Machines replacing humans

Humans and machines working

alongside each other

Machines better

understanding humans and

the environment

Humans better understanding

machines

Machines and humans

becoming smarter


Source: Gartner August 2014

Could affect freight transport


Increasing freight transport demand

http://www.eea.europa.eu/data-and-maps/figures/freight-transport-activity-growth-for-eu-25

EU-25

OpportunitiesDigitalisation Increasing goods volumes

New technology

Political

interest

Quad Aces by fitzsean on Flickr

ICT MaturityApp

"Multi-Touch" by DaveLawler on Flickr (CC-BY)

Strategic Tactical Operational Predictive

Time horizons

We are approaching this boundary

…and we are starting to move past it!

Real-time!

Business processes Infrastructure

Paper based Phone

Papers

Road signsAnalogue

tools

RDS

Monitor fuel

cosnumption

Digitalisation version 0 0.5 1.0 1.5 2.0

E-m

ail

Fax

TMS

-

systems

Excel

Route planning

GPS for n

avigatio

n

Electro

nically

genera

ted

freig

ht docum

ents

Barcodes

RFI

D-t

ags

Simple order handling

Advanced order handling

Open interface

Web

based UI

Platform based

systems

Hardw

are-

oriented

Data collection

systems

(prop

rietary)

Com

munication w

ith

vehicles

E-invoice

Web based

booking

Route optimisation

Th

e so

cia

l web

Open connectivity

Integrated

prognosis

Data collection

systems (open)

Tolling

systems

Webservices with

traffic data

Dyn

amic

ro

utin

g sy

stem

s

Pe

rform

an

ce

Ba

sed

ac

ce

ss

Pe

rfo

rma

nc

e

Ba

sed

ac

ce

ss

Mas

hups

Mul

tiple

dat

a so

urce

s

Pro

be

dat

a

Individual

routin

g

inform

ation

Platooning

PlatooningExceptions handling

Sm

art g

ood

s

Manual

Computers

Software

Functions

Dis

trib

uted

deci

sion

m

akin

g

Goods as bi-

directio

nal

hyperlink

Paper based

CC-BY Per Olof Arnäs, Chalmers

Goods VehicleBarcodes

RFID Sensors

ERP systems TMS systems

E-invoices Cloudbased

services

Order handling Driver support Vehicle economics

RDS-TMC Road taxes Active traffic support

Predictive

maintenance

2014-10-15

Goods

Vehicles

Business processes

Infrastructure

Stra-tegic

Tac-tical

Opera-tional

Pre-dictiveWhat happens

when access to real-time data increases?

not quite clear on the concept by woodleywonderworks on Flickr (CC-BY)

The Action of New York City by Trey Ratcliff on Flickr (CC-BY,NC,SA)

Need for speed

Data collection

Data processingData

exploitation

En la cima! by Alejandro Juárez on Flickr (CC-BY)

3 mountaintops to climb…


3 data types

Mountaintop #1

Collection of data in real-time

Fixed Historical Snapshot


Mountaintop #1

Collection of data in real-time

5 data domainsVehicle CargoDriver Company

Infrastructure/facility

at leas

t…

Length Weight WidthHeight

Capacity + other PBS-criteria

EmissionsFuel consumption

Route

Position Speed

Direction

Weight Origin

Destination Accepted ETA

Temperature + other state variables

Temperature + other state variables

Education/training

Speed (ISA) Rest/break schedule

Traffic behaviour Belt usage

Alco lock history

Schedule status (time to next break etc.)

Contracts/ agreements Previous interactions Backoffice support

Fixed Historical Snapshot

Vehicle

Cargo

Driver

Company

Infrastructure/facility

Map + fixed data layers Traffic history

Current traffic Queue

Availability

DATA MATRIX

Mountaintop #2

Processing of data in real-time


Locals and Tourists #1 (GTWA #2): London by Eric Fischer on Flickr

Mountaintop #2

Processing of data in real-time


Mountaintop #3

Exploiting data in real-time


Connected. 362/365 by AndYaDontStop on Flickr (CC-BY)

Lisa for I/O Keynote by Max Braun on Flickr (CC-BY)

Fulham-Manchester United 24-02-2007 by vuhlser on Flickr (CC-

BY)

Mountaintop #3

Exploiting data in real-time


Boeing-KC-97 Stratotanker by x-ray delta one on Flickr (CC-BY)

smile! by Judy van der Velden (CC-BY,NC,SA)

Speculative shipping

http://www.scdigest.com/ontarget/14-01-21-1.php?cid=7767

http://www.scdigest.com/ontarget/14-01-21-1.php?cid=7767

Speculative shipping Package item(s) as a package for

eventual shipment to a delivery address

Associate unique ID with package

Select destination geographic area for package

Ship package to selected distribution geographic area without completely

specifying delivery address

Orders satisfied by item(s)

received?

Package redirected?

Determine package location

Convey delivery address, package ID to delivery location

Assign delivery address to package

Deliver package to delivery address

Convey indication of new destination geographic area and package ID to

current location

Yes

Yes

No

No

smile! by Judy van der Velden (CC-BY,NC,SA)

CASES (MANY)

CASES (MANY MORE)

Big data in freight transport

!

Film by Foursquare. Google: checkins foursquare

”Fast Up-and-Coming Movers Toward the Peak Are Fueled by Digital Business and Payments”

”…the market has settled into a reasonable set of approaches, and the new technologies and practices are additive to existing solutions” (regarding the decline of Big data on the curve)

Gartner, August 2014


So…

What is Big data?

80 by Phil Dragash on Flickr (CC-BY,NC,SA)

2011 2013 2015

”Big data is an all-encompassing term for any collection of data sets so large and complex that it becomes difficult to process using on-hand data management tools or traditional data processing applications.”

- Wikipedia

Google flights

https://www.google.se/flights/

Jawbone measures sleep interruption during earthquake

https://jawbone.com/blog/napa-earthquake-effect-on-sleep/

Not statistics

Exhausted by Adrian Sampson on Flickr (CC-BY)

just

Not Business

Intelligence

Basingstoke Office Staff Desk "No computer" by John Sheldon on Flickr (CC-BY,NC,SA)

just

http://dashburst.com/infographic/big-data-volume-variety-velocity/

Human resources

Reduction in driver turnover, driver

assignment, using sentiment data

analysis

Real-time capacity availability

Inventory management

Examples of applications in freight (Waller and Fawcett, 2013)

Transportation management

Optimal routing, taking into account weather,

traffic congestion, and driver characteristics

Time of delivery, factoring in weather,

driver characteristics, time of day and date

Forecasting

Waller, M. A. and Fawcett, S. E. (2013), Data Science, Predictive Analytics, and Big Data: A Revolution That Will Transform Supply Chain Design and Management. JOURNAL OF BUSINESS LOGISTICS, 34: 77–84

Manage complex systems

Image from: http://www.as-coa.org/watchlisten/ascoa-visits-rios-operations-center

Predict future events

Avoid unpleasant surprises

http://blog.digital.telefonica.com/?press-release=telefonica-dynamic-insights-launches-smart-steps-in-the-uk

Vizualisation

7Big Data Best Practice Across Industries

Usage of data in order to:Increase Level of TransparencyOptimize ResourceConsumption Improve Process Qualityand Performance

Increase customersloyalty and retentionPerforming precisecustomer segmentationand targetingOptimize customerinteraction and service

Expanding revenuestreams from existingproductsCreating new revenuestreams from entirelynew (data) products

Exploit data for: Capitalize on data by:

New Business Models

Customer Experience

OperationalEfficiency

Use data to: • Increase level of

transparency• Optimize resource

consumption • Improve process quality

and performance

Exploit data to: • Increase customer

loyalty and retention• Perform precise customer

segmentation and targeting • Optimize customer interaction

and service

Capitalize on data by: • Expanding revenue streams

from existing products • Creating new revenue

streams from entirely new (data) products

New Business ModelsCustomer ExperienceOperational Efficiency

Figure 4: Value dimensions for Big Data use cases; Source: DPDHL / Detecon

2.1 Operational Efficiency

For metropolitan police departments, the task of tracking down criminals to preserve public safety can sometimes be tedious. With many siloed information repositories, casework often involves making manual connection of many data points. This takes times and dramatically slows case resolution. Moreover, road policing resources are deployed reactively, making it very difficult to catch criminals in the act. In most cases, it is not possible to resolve these challenges by increasing police staffing, as government budgets are limited.

One authority that is leveraging its various data sources is the New York Police Department (NYPD). By capturing and connecting pieces of crime-related information, it hopes to stay one step ahead of the perpetrators of crime.6 Long before the term Big Data was coined, the NYPD made an effort to break up the compartmentalization of its data ingests (e.g., data from 911 calls, investigation reports, and more). With a single view of all the informa-

tion related to one particular crime, officers achieve a more coherent, real-time picture of their cases. This shift has significantly sped up retrospective analysis and allows the NYPD to take action earlier in tracking down individual criminals.

The steadily decreasing rates of violent crime in New York7 have been attributed not only to this more effective streamlining of the many data items required to perform casework but also to a fundamental change in policing practice.8 By introducing statistical analysis and georaphical mapping of crime spots, the NYPD has been able to create a “bigger picture” to guide resource deployment and patrol practice.

Now the department can recognize crime patterns using computational analysis, and this delivers insights enabling each commanding officer to proactively identify hot spots of criminal activity.

6 “NYPD changes the crime control equation by the way it uses information”, IBM; cf. https://www-01.ibm.com/software/success/cssdb.nsf/CS/JSTS-6PFJAZ7 “Index Crimes By Region”, New York State Division of Criminal Justice Services, May 2013, cf. http://www.criminaljustice.ny.gov/crimnet/ojsa/stats.htm8 “Compstat and Organizational Change in the Lowell Police Department”, Willis et. al., Police Foundation, 2004; cf. http://www.policefoundation.org/

content/compstat-and-organizational-change-lowell-police-department

2.1.1 Utilizing data to predict crime hotspots

DHL 2013: ”Big Data in Logistics”

Domain knowledge critical!

See for instance: Waller, M. A. and Fawcett, S. E. (2013), Data Science, Predictive Analytics, and Big Data: A Revolution

That Will Transform Supply Chain Design and Management. JOURNAL OF BUSINESS LOGISTICS, 34: 77–84

Data scientists - the new superstars

"Data Science Venn Diagram" by Drew Conway - Own work. Licensed under Creative Commons Attribution-Share Alike 3.0 via Wikimedia Commons - http://commons.wikimedia.org/wiki/File:Data_Science_Venn_Diagram.png#mediaviewer/File:Data_Science_Venn_Diagram.png

It’s not business as usual.

Hello Kitty Darth Vader by JD Hancock on Flickr (CC-BY)


This is the internet happening to freight

transport.




transport.

There is no ’usual’ anymore.



Get used to it.


transport.

There is no ’usual’ anymore.


Challenges

The Challenger by Martín Vinacur on Flickr (CC-BY)

Challenges


Company

Challenges


Company

Supply chain, simple

Challenges


Company


Supply chain, complex

Challenges


Company


Supply chain, complex

Eco system

Challenges


Cross-disciplinary

Cross-industries

Cross-borders


Not all ideas age with grace

Someone must do the work



Not everyone will want to adopt new things…

!

REAL-TIME (BIG) DATA IN FREIGHT TRANSPORT - MEETING THE GLOBAL TRENDS

Per Olof Arnäs Chalmers University of Technology @Dr_PO [email protected] about.me/perolofarnas Slides: slideshare.net/poar

Image: www.simonstalenhag.se

Real-time (Big) Data in Freight Transport - Meeting the global trends

Business

pwc megatrends

flickr ccbymegatrend

bunch of pilotsand

bunch of terminals

bunch of planeshiring

american airlines

resource flow

resource demandsource