Post on 13-Oct-2020
Approaches and ToolsApproaches and Tools
Optimisation of Urban Freight Systems by strategic co-operations
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations2
First traffic jam in the Netherlands: 1955First traffic jam in the Netherlands: 1955
Utrecht, 1955: 50.000 → 2006: 350.000 vehicles/day
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations3
Impact of restrictions on costsImpact of restrictions on costs
Gemiddelde kosten per rolcontainer voor Filiaal in het Centrum
5.4
7.5
0
1
2
3
4
5
6
7
8
9
Gemiddelde kostprijs per rolcontainer filiaal inKernwinkelgebied
Kostprijs per rolcontainer exclusiefvenstertijden in het Kernwinkelgebied
Eur
o/R
olco
ntai
ner
Average costs per rolling container forstore in city centre
with restrictions without restrictions
Dutch retail:€ 425 million/yearDutch retail:Dutch retail:€ 425 € 425 million/yearmillion/year
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations4
Regulation per municipality in the NetherlandsRegulation per municipality in the Netherlands
no time windows or vehicle restrictionsvehicle restrictionstime windowstime windows and vehicle restrictions
• over 16 million inhabitants• 180.000 shops/retail outlets• 5300 supermarkets• 53% of cities have Time Windows• 43% have Vehicle Restrictions• 59% have Restricted Access to Center
of which 21% permanent• 6,800,000 cars• 760,000 vans• 165,000 trucks
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations5
Distribution of the deliveries over the day (as percentage of the total number of consignments in 24 hours)
17.7%
1.6% 2.7%
12.6%
0.1%
65.2%
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
70.0%
06:00-11:00 11:00-18:00 18:00-21:00 23:00-06:00 variable unknown
The situation todayThe situation today
• Peak in capacity needed• Partially Self induced by retail• Congestion • Personnel present at stores
Tilburg, Amsterdam, Breda, Utrecht, RotterdamLeiden, Den Haag, OosterhoutGroningen, LeeuwardenDen Bosch
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations6
Recent surveyRecent survey
What is the impact of municipal regulation regarding city What is the impact of municipal regulation regarding city distribution of cargo on vehicle kilometres and fleet size?distribution of cargo on vehicle kilometres and fleet size?
• Different scenarios• Data of:
• 20 chains of supermarkets (2249 st, 35 dc)• 3 chains of department stores (328 st, 5 dc)• 4 chains of drugstores (1777 st, 5 dc)• 4 chains of electronic equipment stores (285 st, 6 dc)
• Per DC and store:• locations, opening hours, volumes, time windows, frequency,
vehicle restrictions, road network, vehicle characteristics
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations7
Stores and DC’sStores and DC’s
Each chain has stores in municipalities of all restriction types!Each chain has stores in municipalities of all restriction typesEach chain has stores in municipalities of all restriction types!! Synergy potential between DCs not used! > separate networksSynergy potential between Synergy potential between DCsDCs not used! > separate networksnot used! > separate networks
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations8
The Hague The Hague
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations9
Integrative modeling approachIntegrative modeling approach
Network designand optimization
Cost calculation Evaluation of thenetwork design
Network design ABC module Simulation module
1 2 3
Validation of the designand robustness
StrategicStructure and alignment
of the networkFacility locationInventory managementTransporation
OperationalScheduling and resource
managementSimulation techniquesVehicle RoutingScheduling
RESPONSETM
Network Design and Optimization Suite
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations10
Model usedModel used
• Vehicle routing problem with time windows (VRPTW)
• To complex for exact solution > heuristic approach
• Based on genetic algorithm of Homberger and Gehring (1999)
• Minimise costs (time or distance) delivering all customers within restrictions
( )( )( )( ) ( )
( ) ( )( )
∑ ∑
∑
= =+
=
⋅
+
−
+⋅+⋅
NofRoutes
r
rrsInRouteNofCustome
cc,rroute,c,rroute
NofRoutes
r depotID,rrsInRouteNofCustome,rroute
distancecpkm
,rleavetimetraveltimerrsInRouteNofCustome,rleavetime
cphNofRoutescpr
1 01
10
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations11
Overview of scenariosOverview of scenarios
NoYesSequential8 Local sequential time windows
NoYesWide7 National uniform wide time window
NoYesYes and No6 Only time windows in big cities
Realistic scenarios:
YesYesYes5 Impact of municipal regulation
YesNoYes4 Impact of time windows
YesYesNo3 Impact vehicle restrictions
NoNoNo2 Impact in night distribution
YesNoNo1 No restrictions
Theoretical scenarios:Theoretical scenarios:
NoYesYesCurrent situation
In night distributionVehicle RestrictionsTime WindowsScenario
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations12
VisualiserVisualiser
RandstadRandstad
Convince chain partners by showing the differences between scenarios in a way they understand!Most of the time, they are not mathematicians!
Convince chain partners by showing the differences between scenarios in a way they understand!Most of the time, they are not mathematicians!
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations13
Results Results (1)(1)
138203Cities > 50.000
135168National uniform time window
140211Regional sequential time window
143182Municipal regulation
111128Time windows
137136Vehicle restrictions
97124No In night distribution
215.000 = 100388 = 100No Restrictions
141218Current situation
KilometresFleet
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations14
Flee
tFl
eet
Kilo
met
res
(*10
00)
Kilo
met
res
(*10
00)
current
current
Results Results (2)(2)
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations15
Results Results (3)(3)
Length time window (hours)
Kilo
met
res
(*10
00)
F
leet
4,74,7
6%!
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations16
Results Results (4)(4)
-39%-63%Electronic equipment stores
-33%-42%Drug stores
-43%-67%Department stores
-20%-53%Supermarkets
KilometresFleetImpact of restrictions
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations17
Results Results (5)(5)
Regional partial sequential time windows: hardly any impact!
• In general not much regional routes but optimal spread of drops
• Optimal sequence different per chain of stores or suppliers
• Trucks used whole day: combination with no-time window areas
• Reason why increase of fleet is caused by trucks, not by vans
• Balance vehicle capacity and optimal trip length (driver shifts)
• Density of time window zones too low
• Only for regional chains of stores in an area with high density of cities with time windows
• Full sequential in bigger area (morning/afternoon)
• Bundling via city distribution centre
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations18
ConclusionsConclusions
• Regional tuning of small time windows gives no improvements
• Impact of municipal regulation mainly depends on length of time windows; a small increase shows already benefits without additional disadvantages
• Small time windows deliver local improvement on liveability but deterioration on national scale due to increase of fleet and kilometres
• Different focus/scale of chains of stores and municipalities cause at the end only negative impact of regulations (one full network versus relevant parts of all networks)
• Bundling of several suppliers in one truck seems to be the only way out
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations19
Time windowsVehicle restrictions
E
Retail warehouse
Fresh/Apparel/FoodNon-Food
Finished product incity box
Outlet in city center
S
Supermarket/outlet/store/consumer
Time windowsVehicle restrictions
An alternative using the CityAn alternative using the City--BoxBox
• Current situation in 53% of Dutch cities
E
Retail warehouse
Fresh/Apparel/FoodNon-Food
Finished productS
Outlet in urban area
S
Supermarket/outlet/store/consumerSmall trucks due to restrictions
Peak in before 11:00 amLoss of capacity during the rest of the day
• Structure using the city-box
Crossdocklocation
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations20
An alternative using the CityAn alternative using the City--BoxBox
E
Retail warehouse
Fresh/Apparel/FoodNon-Food
Finished product incity box
Outlet in city center
S
Supermarket/outlet/store/consumer
Crossdocklocation
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations21
An alternative using the CityAn alternative using the City--BoxBox
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations22
An alternative using the CityAn alternative using the City--BoxBox
• City box standard loading unit• 2.25x2.55x2.25 meters• rolling containers, pallets, boxes and racks• lift-rolling
• minimizing the handling on crossdock• standardization city-box• standardization of vehicles
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations23
VisualiserVisualiser
TilburgTilburg
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations24
Impact of noise / emissionImpact of noise / emission
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations25
Impact of CityImpact of City--Box on kilometres, routes and Box on kilometres, routes and emissionemission
Emissies Tilburg
13
9
0
20
40
60
80
100
120
Huidige situatie Inzet stadsbox optie 1 Inzet stadsbox optie 2 Totaal effectN
O2
emission
Totaal Emissies Tilburg
3
9
0
20
40
60
80
100
120
Huidige situatie Huidige dataset Totaal aantal rittengeschat
Totaal effect
NO
2
routes
Vrachtwagenkilometers Tilburg
54007
62286
0
100,000
200,000
300,000
400,000
500,000
600,000
Huidige situatie Inzet stadsbox optie 1 Inzet stadsbox optie 2 Totaal effect
vrac
htw
agen
kilo
met
ers
per
jaar
kilometres
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations26
ConclusionsConclusions
• Less routes due to increased load factor• Reduction of kilometres mainly on line hauls\• Further improvement:
• Frequency• Drop size• Time windows• Vehicle restrictions• Improvement car technology
Vienna, 21 September 2006Optimisation of Urban Freight Systems by strategic co-operations27
More information?More information?
G.W. (Egbert) Guis MScTNOBU Mobility & Logistics
Egbert.Guis@TNO.nlwww.tno.nl/logistiek+31 15 269 6831