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
Nov 19, 2014
!
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
Source: PWC (google: pwc megatrends 2014)
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
Source: PWC (google: pwc megatrends 2014)
¡Rayos! by José Eugenio Gómez Rodríguez on Flickr (CC-BY,NC,SA)
Megatrend #4
Climate change and
resource scarcity
Source: PWC (google: pwc megatrends 2014)
Megatrend #5
Technological breakthroughs
Source: PWC (google: pwc megatrends 2014)
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
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
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
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
Gartners Hype Cycle for Emerging Technologies
Source: Gartner August 2014
Could affect freight transport
Gartners Hype Cycle for Emerging Technologies
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…
En la cima! by Alejandro Juárez on Flickr (CC-BY)
3 data types
Mountaintop #1
Collection of data in real-time
Fixed Historical Snapshot
En la cima! by Alejandro Juárez on Flickr (CC-BY)
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
En la cima! by Alejandro Juárez on Flickr (CC-BY)
Locals and Tourists #1 (GTWA #2): London by Eric Fischer on Flickr
Mountaintop #2
Processing of data in real-time
En la cima! by Alejandro Juárez on Flickr (CC-BY)
Mountaintop #3
Exploiting data in real-time
En la cima! by Alejandro Juárez on Flickr (CC-BY)
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
En la cima! by Alejandro Juárez on Flickr (CC-BY)
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
Gartners Hype Cycle for Emerging Technologies
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)
It’s not business as usual.
This is the internet happening to freight
transport.
Hello Kitty Darth Vader by JD Hancock on Flickr (CC-BY)
It’s not business as usual.
This is the internet happening to freight
transport.
There is no ’usual’ anymore.
Hello Kitty Darth Vader by JD Hancock on Flickr (CC-BY)
It’s not business as usual.
Get used to it.
This is the internet happening to freight
transport.
There is no ’usual’ anymore.
Hello Kitty Darth Vader by JD Hancock on Flickr (CC-BY)
Challenges
The Challenger by Martín Vinacur on Flickr (CC-BY)
Challenges
The Challenger by Martín Vinacur on Flickr (CC-BY)
Company
Challenges
The Challenger by Martín Vinacur on Flickr (CC-BY)
Company
Supply chain, simple
Challenges
The Challenger by Martín Vinacur on Flickr (CC-BY)
Company
Supply chain, simple
Supply chain, complex
Challenges
The Challenger by Martín Vinacur on Flickr (CC-BY)
Company
Supply chain, simple
Supply chain, complex
Eco system
Challenges
The Challenger by Martín Vinacur on Flickr (CC-BY)
Cross-disciplinary
Cross-industries
Cross-borders
The Challenger by Martín Vinacur on Flickr (CC-BY)
Not all ideas age with grace
Someone must do the work
The Challenger by Martín Vinacur on Flickr (CC-BY)
The Challenger by Martín Vinacur on Flickr (CC-BY)
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