SlideShare a Scribd company logo
1 of 48
Download to read offline
Dynamic	
  Dispatch	
  Waves	
  for	
  
Same-­‐Day	
  Delivery	
  
Mathias	
  Klapp,	
  Alejandro	
  Toriello,	
  
Alan	
  Erera	
  
School	
  of	
  Industrial	
  and	
  Systems	
  Engineering	
  
Georgia	
  Tech	
  	
  
UC-­‐Berkeley	
  ITS	
  Friday	
  Seminar	
  
February	
  20,	
  2015	
  
What	
  to	
  remember	
  
1.  Last-­‐mile	
  home	
  delivery	
  logis=cs	
  costly	
  due	
  
to	
  poor	
  scale	
  economies,	
  and	
  same	
  day	
  
delivery	
  adds	
  to	
  challenge	
  
2.  	
  Dynamic	
  vehicle	
  dispatch	
  strategies	
  for	
  SDD	
  
systems	
  may	
  provide	
  significant	
  value	
  over	
  
fixed	
  wave	
  strategies	
  
3.  Simple	
  rollout	
  policies	
  produce	
  high	
  quality	
  
dynamic	
  solu=ons	
  in	
  idealized	
  seJng	
  	
  
This	
  talk	
  is	
  not	
  about…	
  
Last-­‐mile	
  home	
  delivery	
  
•  Weak	
  scale	
  economies	
  
– Ton-­‐miles	
  /	
  operator-­‐hour	
  compara=vely	
  low	
  
– Small	
  vehicles,	
  opera=ng	
  cost	
  inefficient	
  
Distribution
center
Consumer delivery locations
Home	
  delivery	
  e-­‐commerce	
  
Home	
  delivery	
  e-­‐commerce	
  
Same-­‐day	
  home	
  delivery	
  
Same-­‐day	
  home	
  delivery	
  
Pick/pack/load	
  and	
  vehicle	
  dispatch	
  
•  Both	
  benefit	
  from	
  order	
  batching	
  
– Pick	
  density	
  for	
  warehouse	
  opera=ons	
  
– Stop	
  density	
  for	
  vehicle	
  rou=ng	
  opera=ons	
  
Pick/pack/load	
  and	
  vehicle	
  dispatch	
  
•  Both	
  benefit	
  from	
  order	
  batching	
  
– Pick	
  density	
  for	
  warehouse	
  opera=ons	
  
– Stop	
  density	
  for	
  vehicle	
  rou=ng	
  opera=ons	
  
Distribution
center
dense = shorter travel
time per delivery
Pick/pack/load	
  and	
  vehicle	
  dispatch	
  
•  Both	
  benefit	
  from	
  order	
  batching	
  
– Pick	
  density	
  for	
  warehouse	
  opera=ons	
  
– Stop	
  density	
  for	
  vehicle	
  rou=ng	
  opera=ons	
  
Distribution
center
sparse = longer travel
time per delivery
Next-­‐day	
  vs.	
  same-­‐day	
  	
  
yesterday today time
orders arrive
Next-day Local Distribution System
order pick, pack, and load
vehicles for delivery dispatched
Next-­‐day	
  vs.	
  same-­‐day	
  	
  
yesterday today time
orders arrive
Same-day Local Distribution System
order pick, pack, and load
vehicles for delivery dispatched
Pick/pack/load	
  batching	
  economies?	
  
yesterday today time
orders arrive
Same-day Local Distribution System
order pick, pack, and load
vehicles for delivery dispatched
many orders arrive after
first picks must be made
Dispatch	
  batching	
  economies?	
  
yesterday today time
orders arrive
Same-day Local Distribution System
order pick, pack, and load
vehicles for delivery dispatched
some vehicles should
be dispatched before all
orders are ready
Vehicle	
  dispatch	
  challenges	
  
•  Each	
  vehicle	
  dispatched	
  mul=ple	
  =mes	
  during	
  
opera=ng	
  day	
  (10-­‐12	
  opera=ng	
  hours)	
  
– When	
  to	
  dispatch	
  vehicles?	
  
•  Tradeoffs	
  between	
  wai=ng	
  to	
  dispatch,	
  
dispatching	
  long	
  routes,	
  dispatching	
  short	
  
routes	
  
– When	
  to	
  wait	
  to	
  accumulate	
  stop	
  density?	
  
– Which	
  orders	
  to	
  serve	
  with	
  each	
  vehicle	
  
dispatch?	
  
Dynamic	
  Dispatch	
  Waves	
  Problem	
  
•  Determine	
  dispatch	
  epochs	
  dynamically	
  
•  Explore	
  tradeoffs	
  ini=ally	
  with	
  single	
  vehicle	
  
and	
  simplified	
  geography	
  
time
wait
dispatch 1 dispatch 2 dispatch 3
Simplified	
  geography:	
  stops	
  on	
  line	
  
di
•  Order	
  loca=on	
  	
  
– round-­‐trip	
  travel	
  =me	
  
from	
  DC	
  
•  No	
  stop	
  =me	
  
•  Dispatch	
  	
  
– serves	
  all	
  ready	
  orders	
  
di
di
{j : dj  di}
Order	
  ready	
  Mme	
  process	
  
di
time
Ready is picked, packed
for loading (no duration)
T 0⌧i
Orders served and
vehicle back to DC
by time 0
Order	
  ready	
  Mme	
  process	
  
di
time
Ready orders for first
dispatch of day
T 0⌧i
Order	
  ready	
  Mme	
  process	
  
di
time
Orders that come
available later in
operating day,
and unknown when
planning at time T
T 0⌧i
Dynamic	
  Dispatch	
  Waves	
  Problem	
  
•  Each	
  =me	
  vehicle	
  at	
  distribu=on	
  center,	
  decide:	
  
–  Whether	
  to	
  dispatch	
  vehicle,	
  or	
  wait	
  
–  If	
  dispatched,	
  which	
  unserved	
  ready	
  orders	
  to	
  include	
  
in	
  the	
  route	
  
•  Given	
  set	
  of	
  poten.al	
  orders	
  
–  Round-­‐trip	
  dispatch	
  =me	
  
–  Stochas=c	
  =me	
  (or	
  wave)	
  when	
  order	
  ready	
  
–  Penalty	
  if	
  order	
  remains	
  unserved	
  
•  Operate	
  to	
  minimize	
  total	
  cost	
  of	
  all	
  dispatches	
  
plus	
  total	
  unserved	
  order	
  penal=es	
  
N = {1, ..., n}
di
⌧i
i
Dynamic	
  programming	
  formulaMon	
  
for	
  DDWP	
  on	
  the	
  line	
  
•  State:	
  
–  Number	
  of	
  remaining	
  waves,	
  
–  Ready	
  and	
  unserved	
  orders,	
  	
  
–  Poten=al	
  orders	
  not	
  yet	
  ready,	
  
•  Ac=ons:	
  	
  wait	
  one	
  wave,	
  or	
  serve	
  
–  Cost:	
  
–  Possible	
  dispatch	
  ac=ons:	
  
–  Must	
  return	
  by	
  0:	
  
•  At	
  end	
  horizon,	
  pay	
  penal=es	
  for	
  unserved	
  orders	
  
(t, R, P)
t
R
P
S ✓ R
|R|
x = maxi2S di
x  t
Dynamic	
  programming	
  formulaMon	
  
for	
  DDWP	
  on	
  the	
  line	
  
Bellman	
  recursion	
  for	
  DDWP	
  
DeterminisMc	
  DDWP	
  on	
  line	
  
•  Request	
  ready	
  =mes	
  known	
  in	
  advance,	
  but	
  
requests	
  cannot	
  be	
  served	
  before	
  ready	
  =me	
  
•  Proper=es	
  of	
  op=mal	
  solu=on	
  
–  Dispatch	
  lengths	
  	
  x	
  	
  strictly	
  decreasing	
  
–  No	
  wai=ng	
  a]er	
  first	
  dispatch	
  
	
  
DeterminisMc	
  DDWP	
  on	
  line	
  
•  Request	
  ready	
  =mes	
  known	
  in	
  advance,	
  but	
  
requests	
  cannot	
  be	
  served	
  before	
  ready	
  =me	
  
•  Proper=es	
  of	
  op=mal	
  solu=on	
  
–  Dispatch	
  lengths	
  	
  x	
  	
  strictly	
  decreasing	
  
–  No	
  wai=ng	
  a]er	
  first	
  dispatch	
  
	
  
DeterminisMc	
  DDWP	
  on	
  line	
  
•  New	
  DP	
  state:	
  remaining	
  waves	
  t,	
  length	
  d	
  of	
  prior	
  dispatch	
  
•  Recursion	
   O(n2
T)
Using	
  determinisMc	
  DDWP	
  
•  Es=ma=ng	
  an	
  a	
  posteriori	
  cost	
  lower	
  bound	
  	
  
– Average	
  cost	
  for	
  sample	
  of	
  order	
  realiza.on	
  days	
  
– Any	
  dynamic	
  policy	
  for	
  stochas=c	
  DDWP	
  can	
  have	
  
no	
  lower	
  expected	
  cost	
  
•  Building	
  a	
  priori	
  policy	
  solu=ons	
  to	
  the	
  
stochas=c	
  DDWP	
  
A	
  priori	
  soluMon	
  
•  Before	
  first	
  dispatch	
  (i.e.,	
  at	
  wave	
  T),	
  find	
  	
  
complete	
  set	
  of	
  vehicle	
  dispatches:	
  
•  Theorem	
  
{(xk
, tk
)}
Optimal a priori solution is solution to
deterministic DDWP where each order i
replicated for each wave t 2 {T, ..., 1} with
known ready time t and penalty i Pr(⌧i = t)
Dynamic	
  policies	
  
1.  Implement	
  a	
  priori	
  solu=on,	
  but	
  adjust	
  during	
  
opera=ons	
  
–  Shorten,	
  delay,	
  and	
  cancel	
  some	
  dispatches	
  
2.  Rollout	
  using	
  a	
  priori	
  solu=ons	
  
– Execute	
  first	
  decision	
  in	
  adjusted	
  a	
  priori	
  solu=on	
  
– Build	
  new	
  a	
  priori	
  plan	
  any	
  =me	
  vehicle	
  at	
  
distribu=on	
  center,	
  using	
  new	
  informa=on	
  
Experiment	
  1	
  
•  Request	
  ready	
  =me	
  process	
  
– Condi=onal	
  arrival	
  likelihoods	
  	
  	
  	
  	
  	
  	
  	
  	
  each	
  wave	
  
•  Request	
  loca=ons	
  and	
  penal=es	
  
– Loca=on	
  discrete	
  uniform	
  up	
  to	
  a	
  maximum	
  
– Penal=es	
  discrete	
  uniform	
  on	
  quarters	
  of	
  
•  	
  20	
  random	
  instances	
  for	
  class	
  
– r	
  measures	
  =me	
  flexibility	
  
✓i
T
`
`
✓
n, `, r =
T
`
◆
Experiment	
  1:	
  Results	
  AvgGaptoaposteriorilowerbound
Experiment	
  1:	
  Results	
  AvgGaptoaposteriorilowerbound
Experiment	
  1:	
  Results,	
  r	
  =	
  2	
  AvgGaptoaposteriorilowerbound
Dynamic	
  policies	
  via	
  ALP	
  
•  Dual	
  LP	
  reformula=on	
  of	
  Bellman’s	
  equa=on	
  
– massive	
  LP:	
  exponen=al	
  variables,	
  constraints	
  
max ERT
[CT (RT , N  RT )]
C0(R, P) 
X
i2R
i
Ct(R, P)  EF t
1
[Ct 1(R [ Ft
1, P  Ft
1)]
Ct(R, P)  d + EF t
d
[Ct d(Rd [ Ft
d, P  Ft
d)]
Dynamic	
  policies	
  via	
  ALP	
  
•  ALP	
  restric=on	
  provides	
  lower	
  bound,	
  and	
  
poten=ally	
  useful	
  approxima=on	
  of	
  C	
  
•  Restrict	
  C	
  :	
  
•  “cost	
  of	
  unserved	
  known	
  requests”,	
  “cost	
  of	
  
unserved	
  poten=al	
  requests”,	
  “value	
  of	
  
remaining	
  waves”	
  
Ct(R, P) ⇡
X
i2R
at
i +
X
j2P
bt
j
tX
k=1
vk
Dynamic	
  policies	
  via	
  ALP	
  
•  Proposi=ons	
  
– Using	
  this	
  restric=on	
  in	
  dual	
  LP,	
  the	
  ALP	
  lower	
  
bound	
  LP	
  requires	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  variables	
  and	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
constraints	
  
– (*)	
  For	
  determinis=c	
  problems,	
  the	
  ALP	
  lower	
  
bound	
  is	
  =ght,	
  equal	
  to	
  op=mal	
  cost	
  
•  Hybrid	
  ALP-­‐A	
  priori	
  rollout	
  policy	
  
– Use	
  a	
  priori	
  rollout	
  first,	
  then	
  switch	
  to	
  ALP	
  rollout	
  
later	
  in	
  opera=ng	
  period	
  
O(nT) O(n2
T)
Experiment	
  1:	
  Results	
  AvgGaptoaposteriorilowerbound
Experiment	
  2	
  
•  New	
  ready	
  =me	
  process	
  
– p:	
  Likelihood	
  ready	
  by	
  T	
  
– q:	
  Likelihood	
  of	
  no	
  request	
  
– Remaining	
  likelihood	
  discrete	
  uniform:	
  
•  Request	
  loca=ons	
  and	
  penal=es	
  as	
  before	
  
•  20	
  random	
  instances	
  
	
  
(n = 20, ` = 10, r = 3)
µi
µi + vµi v
Experiment	
  2:	
  Results	
  vs.	
  q	
  -­‐	
  p	
  AvgGaptoaposteriorilowerbound
Experiment	
  2:	
  Results	
  vs.	
  v	
  AvgGaptoaposteriorilowerbound
ObservaMons:	
  1	
  
•  Dynamic	
  soluMons	
  valuable	
  
– Dispatching	
  scheme	
  from	
  A	
  priori-­‐rollout	
  
approach	
  usually	
  provides	
  significant	
  savings	
  over	
  
instance-­‐specific	
  A	
  priori	
  solu=ons	
  
– Schemes	
  with	
  fixed	
  dispatch	
  waves	
  could	
  be	
  no	
  
befer	
  in	
  this	
  seJng	
  
time
wave I wave II wave III
ObservaMons:	
  1	
  
•  Fixed-­‐but-­‐flexible	
  dispatch	
  waves	
  (?)	
  
– Fixed	
  planning	
  waves	
  useful	
  to	
  DC	
  pick/pack/load,	
  
and	
  for	
  customer	
  order	
  management	
  
– Design	
  and	
  performance	
  of	
  a	
  fixed-­‐but-­‐flexible	
  
dispatch	
  wave	
  system?	
  
time
wave I wave II wave III
ObservaMons:	
  2	
  
•  LocaMons	
  on	
  line	
  creates	
  maximum	
  batching	
  
benefit	
  
– Compounded	
  by	
  assump=on	
  of	
  no	
  fixed	
  stop	
  =me	
  
required	
  per	
  delivery	
  
– Incen=ve	
  to	
  wait	
  and	
  batch	
  may	
  be	
  too	
  strong	
  
– Inves=ga=ng	
  problems	
  with	
  fixed	
  stop	
  =mes	
  and	
  
two-­‐dimensional	
  delivery	
  loca=ons	
  
	
  
ObservaMons:	
  2	
  
T 0⌧i
latest wave and
dispatch duration
Other	
  extensions	
  
•  MulMple	
  vehicles	
  per	
  delivery	
  zone	
  
– How	
  to	
  coordinate	
  dispatch	
  waves	
  for	
  two	
  
vehicles	
  serving	
  a	
  single	
  zone?	
  	
  Other	
  
configura=ons?	
  
•  Customer	
  order	
  management	
  
– Reject/not	
  offer	
  same	
  day	
  delivery	
  op=on	
  
dynamically	
  as	
  orders	
  are	
  received	
  
What	
  to	
  remember	
  
1.  Last-­‐mile	
  home	
  delivery	
  logis=cs	
  costly	
  due	
  
to	
  poor	
  scale	
  economies,	
  and	
  same	
  day	
  
delivery	
  adds	
  to	
  challenge	
  
2.  	
  Dynamic	
  vehicle	
  dispatch	
  strategies	
  for	
  SDD	
  
systems	
  may	
  provide	
  significant	
  value	
  over	
  
fixed	
  wave	
  strategies	
  
3.  Simple	
  rollout	
  policies	
  produce	
  high	
  quality	
  
dynamic	
  solu=ons	
  in	
  idealized	
  seJng	
  	
  

More Related Content

What's hot

Mpage Scicomp14 Bynd Tsk Geom
Mpage Scicomp14 Bynd Tsk GeomMpage Scicomp14 Bynd Tsk Geom
Mpage Scicomp14 Bynd Tsk Geommpage_colo
 
Signalized Intersections (Transportation Engineering)
Signalized Intersections (Transportation Engineering)Signalized Intersections (Transportation Engineering)
Signalized Intersections (Transportation Engineering)Hossam Shafiq I
 
Lec 13A Signalized Intersections (Transportation Engineering Dr.Lina Shbeeb)
Lec 13A Signalized Intersections (Transportation Engineering Dr.Lina Shbeeb)Lec 13A Signalized Intersections (Transportation Engineering Dr.Lina Shbeeb)
Lec 13A Signalized Intersections (Transportation Engineering Dr.Lina Shbeeb)Hossam Shafiq I
 
Parallel Programming for Multi- Core and Cluster Systems - Performance Analysis
Parallel Programming for Multi- Core and Cluster Systems - Performance AnalysisParallel Programming for Multi- Core and Cluster Systems - Performance Analysis
Parallel Programming for Multi- Core and Cluster Systems - Performance AnalysisShah Zaib
 
Unit7 & 8 Performance and optimization
Unit7 & 8 Performance and optimization Unit7 & 8 Performance and optimization
Unit7 & 8 Performance and optimization leenachandra
 

What's hot (6)

Mpage Scicomp14 Bynd Tsk Geom
Mpage Scicomp14 Bynd Tsk GeomMpage Scicomp14 Bynd Tsk Geom
Mpage Scicomp14 Bynd Tsk Geom
 
L16 Progression
L16 ProgressionL16 Progression
L16 Progression
 
Signalized Intersections (Transportation Engineering)
Signalized Intersections (Transportation Engineering)Signalized Intersections (Transportation Engineering)
Signalized Intersections (Transportation Engineering)
 
Lec 13A Signalized Intersections (Transportation Engineering Dr.Lina Shbeeb)
Lec 13A Signalized Intersections (Transportation Engineering Dr.Lina Shbeeb)Lec 13A Signalized Intersections (Transportation Engineering Dr.Lina Shbeeb)
Lec 13A Signalized Intersections (Transportation Engineering Dr.Lina Shbeeb)
 
Parallel Programming for Multi- Core and Cluster Systems - Performance Analysis
Parallel Programming for Multi- Core and Cluster Systems - Performance AnalysisParallel Programming for Multi- Core and Cluster Systems - Performance Analysis
Parallel Programming for Multi- Core and Cluster Systems - Performance Analysis
 
Unit7 & 8 Performance and optimization
Unit7 & 8 Performance and optimization Unit7 & 8 Performance and optimization
Unit7 & 8 Performance and optimization
 

Viewers also liked

Freight Logistics Fundamentals
Freight Logistics FundamentalsFreight Logistics Fundamentals
Freight Logistics FundamentalsAlan Erera
 
Supply Chain Metrics That Matter: Third Party Logistics Providers-10 DEC 2013
Supply Chain Metrics That Matter: Third Party Logistics Providers-10 DEC 2013Supply Chain Metrics That Matter: Third Party Logistics Providers-10 DEC 2013
Supply Chain Metrics That Matter: Third Party Logistics Providers-10 DEC 2013Lora Cecere
 
Logistics Collaboration Solutions-Research
Logistics Collaboration Solutions-ResearchLogistics Collaboration Solutions-Research
Logistics Collaboration Solutions-ResearchAmit Kumar Garg
 
Logistics and supply chain May 2015
Logistics and supply chain May 2015Logistics and supply chain May 2015
Logistics and supply chain May 2015Shipping English
 
Logistics Transportation
Logistics TransportationLogistics Transportation
Logistics Transportationshri1984
 
Overview of Southern Fried Supernova
Overview of Southern Fried SupernovaOverview of Southern Fried Supernova
Overview of Southern Fried SupernovaHypepotamus
 
Ecommerce School: Roger Lopez on PPC on a Shoestring Budget
Ecommerce School: Roger Lopez on PPC on a Shoestring BudgetEcommerce School: Roger Lopez on PPC on a Shoestring Budget
Ecommerce School: Roger Lopez on PPC on a Shoestring BudgetHypepotamus
 
Ecommerce School: Roger Lopez on
Ecommerce School: Roger Lopez on Ecommerce School: Roger Lopez on
Ecommerce School: Roger Lopez on Hypepotamus
 
Ecommerce School: Roger Lopez on Email Marketing 101
Ecommerce School: Roger Lopez on Email Marketing 101Ecommerce School: Roger Lopez on Email Marketing 101
Ecommerce School: Roger Lopez on Email Marketing 101Hypepotamus
 
Ecommerce School: Roger Lopez on Tips on Migrating Ecommerce Platform
Ecommerce School: Roger Lopez on Tips on Migrating Ecommerce PlatformEcommerce School: Roger Lopez on Tips on Migrating Ecommerce Platform
Ecommerce School: Roger Lopez on Tips on Migrating Ecommerce PlatformHypepotamus
 
E commerce marketing tricks
E commerce marketing tricksE commerce marketing tricks
E commerce marketing tricksHypepotamus
 
Summary of Our First Six Months
Summary of Our First Six MonthsSummary of Our First Six Months
Summary of Our First Six MonthsHypepotamus
 
Riding the Cloud Leveraging your Carrier Software to Streamline Pick/Pack/Ship
Riding the Cloud Leveraging your Carrier Software to Streamline Pick/Pack/ShipRiding the Cloud Leveraging your Carrier Software to Streamline Pick/Pack/Ship
Riding the Cloud Leveraging your Carrier Software to Streamline Pick/Pack/ShipOz Development
 
Summer Startup Academy Session 4 with Nick Park from Autotrader.com
Summer Startup Academy Session 4 with Nick Park from Autotrader.comSummer Startup Academy Session 4 with Nick Park from Autotrader.com
Summer Startup Academy Session 4 with Nick Park from Autotrader.comHypepotamus
 
Ecommerce School: Roger Lopez on Google Analytics
Ecommerce School: Roger Lopez on Google AnalyticsEcommerce School: Roger Lopez on Google Analytics
Ecommerce School: Roger Lopez on Google AnalyticsHypepotamus
 
Ecommerce School: Roger Lopez on How to Select an Ecommerce Platform
Ecommerce School: Roger Lopez on How to Select an Ecommerce PlatformEcommerce School: Roger Lopez on How to Select an Ecommerce Platform
Ecommerce School: Roger Lopez on How to Select an Ecommerce PlatformHypepotamus
 
Ecommerce School: Roger Lopez on Content Strategy Framework for Ecommerce Web...
Ecommerce School: Roger Lopez on Content Strategy Framework for Ecommerce Web...Ecommerce School: Roger Lopez on Content Strategy Framework for Ecommerce Web...
Ecommerce School: Roger Lopez on Content Strategy Framework for Ecommerce Web...Hypepotamus
 
Greece : A sea & land logistical hub through the centuries turns to be the ga...
Greece : A sea & land logistical hub through the centuries turns to be the ga...Greece : A sea & land logistical hub through the centuries turns to be the ga...
Greece : A sea & land logistical hub through the centuries turns to be the ga...Antonis Antoniadis
 
Ecommerce School: Roger Lopez on Digital Measurement Model
Ecommerce School: Roger Lopez on Digital Measurement ModelEcommerce School: Roger Lopez on Digital Measurement Model
Ecommerce School: Roger Lopez on Digital Measurement ModelHypepotamus
 

Viewers also liked (20)

Freight Logistics Fundamentals
Freight Logistics FundamentalsFreight Logistics Fundamentals
Freight Logistics Fundamentals
 
Supply Chain Metrics That Matter: Third Party Logistics Providers-10 DEC 2013
Supply Chain Metrics That Matter: Third Party Logistics Providers-10 DEC 2013Supply Chain Metrics That Matter: Third Party Logistics Providers-10 DEC 2013
Supply Chain Metrics That Matter: Third Party Logistics Providers-10 DEC 2013
 
XPO_Group_Presentation_en
XPO_Group_Presentation_enXPO_Group_Presentation_en
XPO_Group_Presentation_en
 
Logistics Collaboration Solutions-Research
Logistics Collaboration Solutions-ResearchLogistics Collaboration Solutions-Research
Logistics Collaboration Solutions-Research
 
Logistics and supply chain May 2015
Logistics and supply chain May 2015Logistics and supply chain May 2015
Logistics and supply chain May 2015
 
Logistics Transportation
Logistics TransportationLogistics Transportation
Logistics Transportation
 
Overview of Southern Fried Supernova
Overview of Southern Fried SupernovaOverview of Southern Fried Supernova
Overview of Southern Fried Supernova
 
Ecommerce School: Roger Lopez on PPC on a Shoestring Budget
Ecommerce School: Roger Lopez on PPC on a Shoestring BudgetEcommerce School: Roger Lopez on PPC on a Shoestring Budget
Ecommerce School: Roger Lopez on PPC on a Shoestring Budget
 
Ecommerce School: Roger Lopez on
Ecommerce School: Roger Lopez on Ecommerce School: Roger Lopez on
Ecommerce School: Roger Lopez on
 
Ecommerce School: Roger Lopez on Email Marketing 101
Ecommerce School: Roger Lopez on Email Marketing 101Ecommerce School: Roger Lopez on Email Marketing 101
Ecommerce School: Roger Lopez on Email Marketing 101
 
Ecommerce School: Roger Lopez on Tips on Migrating Ecommerce Platform
Ecommerce School: Roger Lopez on Tips on Migrating Ecommerce PlatformEcommerce School: Roger Lopez on Tips on Migrating Ecommerce Platform
Ecommerce School: Roger Lopez on Tips on Migrating Ecommerce Platform
 
E commerce marketing tricks
E commerce marketing tricksE commerce marketing tricks
E commerce marketing tricks
 
Summary of Our First Six Months
Summary of Our First Six MonthsSummary of Our First Six Months
Summary of Our First Six Months
 
Riding the Cloud Leveraging your Carrier Software to Streamline Pick/Pack/Ship
Riding the Cloud Leveraging your Carrier Software to Streamline Pick/Pack/ShipRiding the Cloud Leveraging your Carrier Software to Streamline Pick/Pack/Ship
Riding the Cloud Leveraging your Carrier Software to Streamline Pick/Pack/Ship
 
Summer Startup Academy Session 4 with Nick Park from Autotrader.com
Summer Startup Academy Session 4 with Nick Park from Autotrader.comSummer Startup Academy Session 4 with Nick Park from Autotrader.com
Summer Startup Academy Session 4 with Nick Park from Autotrader.com
 
Ecommerce School: Roger Lopez on Google Analytics
Ecommerce School: Roger Lopez on Google AnalyticsEcommerce School: Roger Lopez on Google Analytics
Ecommerce School: Roger Lopez on Google Analytics
 
Ecommerce School: Roger Lopez on How to Select an Ecommerce Platform
Ecommerce School: Roger Lopez on How to Select an Ecommerce PlatformEcommerce School: Roger Lopez on How to Select an Ecommerce Platform
Ecommerce School: Roger Lopez on How to Select an Ecommerce Platform
 
Ecommerce School: Roger Lopez on Content Strategy Framework for Ecommerce Web...
Ecommerce School: Roger Lopez on Content Strategy Framework for Ecommerce Web...Ecommerce School: Roger Lopez on Content Strategy Framework for Ecommerce Web...
Ecommerce School: Roger Lopez on Content Strategy Framework for Ecommerce Web...
 
Greece : A sea & land logistical hub through the centuries turns to be the ga...
Greece : A sea & land logistical hub through the centuries turns to be the ga...Greece : A sea & land logistical hub through the centuries turns to be the ga...
Greece : A sea & land logistical hub through the centuries turns to be the ga...
 
Ecommerce School: Roger Lopez on Digital Measurement Model
Ecommerce School: Roger Lopez on Digital Measurement ModelEcommerce School: Roger Lopez on Digital Measurement Model
Ecommerce School: Roger Lopez on Digital Measurement Model
 

Similar to Dynamic Dispatch Waves for Same-day Delivery

A Dynamic Logistic Dispatching System With Set-Based Particle Swarm Optimization
A Dynamic Logistic Dispatching System With Set-Based Particle Swarm OptimizationA Dynamic Logistic Dispatching System With Set-Based Particle Swarm Optimization
A Dynamic Logistic Dispatching System With Set-Based Particle Swarm OptimizationRajib Roy
 
Operations Research and Mathematical Modeling
Operations Research and Mathematical ModelingOperations Research and Mathematical Modeling
Operations Research and Mathematical ModelingVinodh Soundarajan
 
1907555 ant colony optimization for simulated dynamic multi-objective railway...
1907555 ant colony optimization for simulated dynamic multi-objective railway...1907555 ant colony optimization for simulated dynamic multi-objective railway...
1907555 ant colony optimization for simulated dynamic multi-objective railway...Mamun Hasan
 
Review of scheduling algorithms in Open Pit Mining
Review of scheduling algorithms in Open Pit MiningReview of scheduling algorithms in Open Pit Mining
Review of scheduling algorithms in Open Pit MiningJose Gonzales, MBA
 
SPLT Transformer.pptx
SPLT Transformer.pptxSPLT Transformer.pptx
SPLT Transformer.pptxSeungeon Baek
 
4. Joe Kraft, Minemax - Minemax Scheduler Applications
4. Joe Kraft, Minemax - Minemax Scheduler Applications4. Joe Kraft, Minemax - Minemax Scheduler Applications
4. Joe Kraft, Minemax - Minemax Scheduler ApplicationsKristy Marshall
 
Reinforcement Learning and Artificial Neural Nets
Reinforcement Learning and Artificial Neural NetsReinforcement Learning and Artificial Neural Nets
Reinforcement Learning and Artificial Neural NetsPierre de Lacaze
 
Queuing model
Queuing model Queuing model
Queuing model goyalrama
 
Dear - 딥러닝 논문읽기 모임 김창연님
Dear - 딥러닝 논문읽기 모임 김창연님Dear - 딥러닝 논문읽기 모임 김창연님
Dear - 딥러닝 논문읽기 모임 김창연님taeseon ryu
 
Ordina Planning & Scheduling Day - APS - Tactical engineering solution - exec...
Ordina Planning & Scheduling Day - APS - Tactical engineering solution - exec...Ordina Planning & Scheduling Day - APS - Tactical engineering solution - exec...
Ordina Planning & Scheduling Day - APS - Tactical engineering solution - exec...Ordina
 
Sara Afshar: Scheduling and Resource Sharing in Multiprocessor Real-Time Systems
Sara Afshar: Scheduling and Resource Sharing in Multiprocessor Real-Time SystemsSara Afshar: Scheduling and Resource Sharing in Multiprocessor Real-Time Systems
Sara Afshar: Scheduling and Resource Sharing in Multiprocessor Real-Time Systemsknowdiff
 
A Lightweight Infrastructure for Graph Analytics
A Lightweight Infrastructure for Graph AnalyticsA Lightweight Infrastructure for Graph Analytics
A Lightweight Infrastructure for Graph AnalyticsDonald Nguyen
 
Design and Analysis of Algorithms.pptx
Design and Analysis of Algorithms.pptxDesign and Analysis of Algorithms.pptx
Design and Analysis of Algorithms.pptxSyed Zaid Irshad
 
Dynamic demand
Dynamic demandDynamic demand
Dynamic demandJumpingJaq
 
Unit7 & 8 performance analysis and optimization
Unit7 & 8 performance analysis and optimizationUnit7 & 8 performance analysis and optimization
Unit7 & 8 performance analysis and optimizationleenachandra
 
Testing Dynamic Behavior in Executable Software Models - Making Cyber-physica...
Testing Dynamic Behavior in Executable Software Models - Making Cyber-physica...Testing Dynamic Behavior in Executable Software Models - Making Cyber-physica...
Testing Dynamic Behavior in Executable Software Models - Making Cyber-physica...Lionel Briand
 
Xilinx timing closure
Xilinx timing closureXilinx timing closure
Xilinx timing closureSampath Reddy
 

Similar to Dynamic Dispatch Waves for Same-day Delivery (20)

A Dynamic Logistic Dispatching System With Set-Based Particle Swarm Optimization
A Dynamic Logistic Dispatching System With Set-Based Particle Swarm OptimizationA Dynamic Logistic Dispatching System With Set-Based Particle Swarm Optimization
A Dynamic Logistic Dispatching System With Set-Based Particle Swarm Optimization
 
Operations Research and Mathematical Modeling
Operations Research and Mathematical ModelingOperations Research and Mathematical Modeling
Operations Research and Mathematical Modeling
 
1907555 ant colony optimization for simulated dynamic multi-objective railway...
1907555 ant colony optimization for simulated dynamic multi-objective railway...1907555 ant colony optimization for simulated dynamic multi-objective railway...
1907555 ant colony optimization for simulated dynamic multi-objective railway...
 
Review of scheduling algorithms in Open Pit Mining
Review of scheduling algorithms in Open Pit MiningReview of scheduling algorithms in Open Pit Mining
Review of scheduling algorithms in Open Pit Mining
 
SPLT Transformer.pptx
SPLT Transformer.pptxSPLT Transformer.pptx
SPLT Transformer.pptx
 
4. Joe Kraft, Minemax - Minemax Scheduler Applications
4. Joe Kraft, Minemax - Minemax Scheduler Applications4. Joe Kraft, Minemax - Minemax Scheduler Applications
4. Joe Kraft, Minemax - Minemax Scheduler Applications
 
Start MPC
Start MPC Start MPC
Start MPC
 
Reinforcement Learning and Artificial Neural Nets
Reinforcement Learning and Artificial Neural NetsReinforcement Learning and Artificial Neural Nets
Reinforcement Learning and Artificial Neural Nets
 
Queuing model
Queuing model Queuing model
Queuing model
 
Dear - 딥러닝 논문읽기 모임 김창연님
Dear - 딥러닝 논문읽기 모임 김창연님Dear - 딥러닝 논문읽기 모임 김창연님
Dear - 딥러닝 논문읽기 모임 김창연님
 
Ordina Planning & Scheduling Day - APS - Tactical engineering solution - exec...
Ordina Planning & Scheduling Day - APS - Tactical engineering solution - exec...Ordina Planning & Scheduling Day - APS - Tactical engineering solution - exec...
Ordina Planning & Scheduling Day - APS - Tactical engineering solution - exec...
 
Sara Afshar: Scheduling and Resource Sharing in Multiprocessor Real-Time Systems
Sara Afshar: Scheduling and Resource Sharing in Multiprocessor Real-Time SystemsSara Afshar: Scheduling and Resource Sharing in Multiprocessor Real-Time Systems
Sara Afshar: Scheduling and Resource Sharing in Multiprocessor Real-Time Systems
 
Os2
Os2Os2
Os2
 
A Lightweight Infrastructure for Graph Analytics
A Lightweight Infrastructure for Graph AnalyticsA Lightweight Infrastructure for Graph Analytics
A Lightweight Infrastructure for Graph Analytics
 
Design and Analysis of Algorithms.pptx
Design and Analysis of Algorithms.pptxDesign and Analysis of Algorithms.pptx
Design and Analysis of Algorithms.pptx
 
Real Time System
Real Time SystemReal Time System
Real Time System
 
Dynamic demand
Dynamic demandDynamic demand
Dynamic demand
 
Unit7 & 8 performance analysis and optimization
Unit7 & 8 performance analysis and optimizationUnit7 & 8 performance analysis and optimization
Unit7 & 8 performance analysis and optimization
 
Testing Dynamic Behavior in Executable Software Models - Making Cyber-physica...
Testing Dynamic Behavior in Executable Software Models - Making Cyber-physica...Testing Dynamic Behavior in Executable Software Models - Making Cyber-physica...
Testing Dynamic Behavior in Executable Software Models - Making Cyber-physica...
 
Xilinx timing closure
Xilinx timing closureXilinx timing closure
Xilinx timing closure
 

Recently uploaded

The Most Attractive Pune Call Girls Budhwar Peth 8250192130 Will You Miss Thi...
The Most Attractive Pune Call Girls Budhwar Peth 8250192130 Will You Miss Thi...The Most Attractive Pune Call Girls Budhwar Peth 8250192130 Will You Miss Thi...
The Most Attractive Pune Call Girls Budhwar Peth 8250192130 Will You Miss Thi...ranjana rawat
 
KubeKraft presentation @CloudNativeHooghly
KubeKraft presentation @CloudNativeHooghlyKubeKraft presentation @CloudNativeHooghly
KubeKraft presentation @CloudNativeHooghlysanyuktamishra911
 
CCS335 _ Neural Networks and Deep Learning Laboratory_Lab Complete Record
CCS335 _ Neural Networks and Deep Learning Laboratory_Lab Complete RecordCCS335 _ Neural Networks and Deep Learning Laboratory_Lab Complete Record
CCS335 _ Neural Networks and Deep Learning Laboratory_Lab Complete RecordAsst.prof M.Gokilavani
 
UNIT-II FMM-Flow Through Circular Conduits
UNIT-II FMM-Flow Through Circular ConduitsUNIT-II FMM-Flow Through Circular Conduits
UNIT-II FMM-Flow Through Circular Conduitsrknatarajan
 
Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...
Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...
Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...roncy bisnoi
 
College Call Girls Nashik Nehal 7001305949 Independent Escort Service Nashik
College Call Girls Nashik Nehal 7001305949 Independent Escort Service NashikCollege Call Girls Nashik Nehal 7001305949 Independent Escort Service Nashik
College Call Girls Nashik Nehal 7001305949 Independent Escort Service NashikCall Girls in Nagpur High Profile
 
Call Girls in Nagpur Suman Call 7001035870 Meet With Nagpur Escorts
Call Girls in Nagpur Suman Call 7001035870 Meet With Nagpur EscortsCall Girls in Nagpur Suman Call 7001035870 Meet With Nagpur Escorts
Call Girls in Nagpur Suman Call 7001035870 Meet With Nagpur EscortsCall Girls in Nagpur High Profile
 
AKTU Computer Networks notes --- Unit 3.pdf
AKTU Computer Networks notes ---  Unit 3.pdfAKTU Computer Networks notes ---  Unit 3.pdf
AKTU Computer Networks notes --- Unit 3.pdfankushspencer015
 
Coefficient of Thermal Expansion and their Importance.pptx
Coefficient of Thermal Expansion and their Importance.pptxCoefficient of Thermal Expansion and their Importance.pptx
Coefficient of Thermal Expansion and their Importance.pptxAsutosh Ranjan
 
University management System project report..pdf
University management System project report..pdfUniversity management System project report..pdf
University management System project report..pdfKamal Acharya
 
Call for Papers - African Journal of Biological Sciences, E-ISSN: 2663-2187, ...
Call for Papers - African Journal of Biological Sciences, E-ISSN: 2663-2187, ...Call for Papers - African Journal of Biological Sciences, E-ISSN: 2663-2187, ...
Call for Papers - African Journal of Biological Sciences, E-ISSN: 2663-2187, ...Christo Ananth
 
ONLINE FOOD ORDER SYSTEM PROJECT REPORT.pdf
ONLINE FOOD ORDER SYSTEM PROJECT REPORT.pdfONLINE FOOD ORDER SYSTEM PROJECT REPORT.pdf
ONLINE FOOD ORDER SYSTEM PROJECT REPORT.pdfKamal Acharya
 
Russian Call Girls in Nagpur Grishma Call 7001035870 Meet With Nagpur Escorts
Russian Call Girls in Nagpur Grishma Call 7001035870 Meet With Nagpur EscortsRussian Call Girls in Nagpur Grishma Call 7001035870 Meet With Nagpur Escorts
Russian Call Girls in Nagpur Grishma Call 7001035870 Meet With Nagpur EscortsCall Girls in Nagpur High Profile
 
Extrusion Processes and Their Limitations
Extrusion Processes and Their LimitationsExtrusion Processes and Their Limitations
Extrusion Processes and Their Limitations120cr0395
 
High Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur EscortsHigh Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur Escortsranjana rawat
 
UNIT-III FMM. DIMENSIONAL ANALYSIS
UNIT-III FMM.        DIMENSIONAL ANALYSISUNIT-III FMM.        DIMENSIONAL ANALYSIS
UNIT-III FMM. DIMENSIONAL ANALYSISrknatarajan
 
Glass Ceramics: Processing and Properties
Glass Ceramics: Processing and PropertiesGlass Ceramics: Processing and Properties
Glass Ceramics: Processing and PropertiesPrabhanshu Chaturvedi
 
VIP Call Girls Ankleshwar 7001035870 Whatsapp Number, 24/07 Booking
VIP Call Girls Ankleshwar 7001035870 Whatsapp Number, 24/07 BookingVIP Call Girls Ankleshwar 7001035870 Whatsapp Number, 24/07 Booking
VIP Call Girls Ankleshwar 7001035870 Whatsapp Number, 24/07 Bookingdharasingh5698
 
BSides Seattle 2024 - Stopping Ethan Hunt From Taking Your Data.pptx
BSides Seattle 2024 - Stopping Ethan Hunt From Taking Your Data.pptxBSides Seattle 2024 - Stopping Ethan Hunt From Taking Your Data.pptx
BSides Seattle 2024 - Stopping Ethan Hunt From Taking Your Data.pptxfenichawla
 
(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...ranjana rawat
 

Recently uploaded (20)

The Most Attractive Pune Call Girls Budhwar Peth 8250192130 Will You Miss Thi...
The Most Attractive Pune Call Girls Budhwar Peth 8250192130 Will You Miss Thi...The Most Attractive Pune Call Girls Budhwar Peth 8250192130 Will You Miss Thi...
The Most Attractive Pune Call Girls Budhwar Peth 8250192130 Will You Miss Thi...
 
KubeKraft presentation @CloudNativeHooghly
KubeKraft presentation @CloudNativeHooghlyKubeKraft presentation @CloudNativeHooghly
KubeKraft presentation @CloudNativeHooghly
 
CCS335 _ Neural Networks and Deep Learning Laboratory_Lab Complete Record
CCS335 _ Neural Networks and Deep Learning Laboratory_Lab Complete RecordCCS335 _ Neural Networks and Deep Learning Laboratory_Lab Complete Record
CCS335 _ Neural Networks and Deep Learning Laboratory_Lab Complete Record
 
UNIT-II FMM-Flow Through Circular Conduits
UNIT-II FMM-Flow Through Circular ConduitsUNIT-II FMM-Flow Through Circular Conduits
UNIT-II FMM-Flow Through Circular Conduits
 
Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...
Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...
Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...
 
College Call Girls Nashik Nehal 7001305949 Independent Escort Service Nashik
College Call Girls Nashik Nehal 7001305949 Independent Escort Service NashikCollege Call Girls Nashik Nehal 7001305949 Independent Escort Service Nashik
College Call Girls Nashik Nehal 7001305949 Independent Escort Service Nashik
 
Call Girls in Nagpur Suman Call 7001035870 Meet With Nagpur Escorts
Call Girls in Nagpur Suman Call 7001035870 Meet With Nagpur EscortsCall Girls in Nagpur Suman Call 7001035870 Meet With Nagpur Escorts
Call Girls in Nagpur Suman Call 7001035870 Meet With Nagpur Escorts
 
AKTU Computer Networks notes --- Unit 3.pdf
AKTU Computer Networks notes ---  Unit 3.pdfAKTU Computer Networks notes ---  Unit 3.pdf
AKTU Computer Networks notes --- Unit 3.pdf
 
Coefficient of Thermal Expansion and their Importance.pptx
Coefficient of Thermal Expansion and their Importance.pptxCoefficient of Thermal Expansion and their Importance.pptx
Coefficient of Thermal Expansion and their Importance.pptx
 
University management System project report..pdf
University management System project report..pdfUniversity management System project report..pdf
University management System project report..pdf
 
Call for Papers - African Journal of Biological Sciences, E-ISSN: 2663-2187, ...
Call for Papers - African Journal of Biological Sciences, E-ISSN: 2663-2187, ...Call for Papers - African Journal of Biological Sciences, E-ISSN: 2663-2187, ...
Call for Papers - African Journal of Biological Sciences, E-ISSN: 2663-2187, ...
 
ONLINE FOOD ORDER SYSTEM PROJECT REPORT.pdf
ONLINE FOOD ORDER SYSTEM PROJECT REPORT.pdfONLINE FOOD ORDER SYSTEM PROJECT REPORT.pdf
ONLINE FOOD ORDER SYSTEM PROJECT REPORT.pdf
 
Russian Call Girls in Nagpur Grishma Call 7001035870 Meet With Nagpur Escorts
Russian Call Girls in Nagpur Grishma Call 7001035870 Meet With Nagpur EscortsRussian Call Girls in Nagpur Grishma Call 7001035870 Meet With Nagpur Escorts
Russian Call Girls in Nagpur Grishma Call 7001035870 Meet With Nagpur Escorts
 
Extrusion Processes and Their Limitations
Extrusion Processes and Their LimitationsExtrusion Processes and Their Limitations
Extrusion Processes and Their Limitations
 
High Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur EscortsHigh Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur Escorts
High Profile Call Girls Nagpur Isha Call 7001035870 Meet With Nagpur Escorts
 
UNIT-III FMM. DIMENSIONAL ANALYSIS
UNIT-III FMM.        DIMENSIONAL ANALYSISUNIT-III FMM.        DIMENSIONAL ANALYSIS
UNIT-III FMM. DIMENSIONAL ANALYSIS
 
Glass Ceramics: Processing and Properties
Glass Ceramics: Processing and PropertiesGlass Ceramics: Processing and Properties
Glass Ceramics: Processing and Properties
 
VIP Call Girls Ankleshwar 7001035870 Whatsapp Number, 24/07 Booking
VIP Call Girls Ankleshwar 7001035870 Whatsapp Number, 24/07 BookingVIP Call Girls Ankleshwar 7001035870 Whatsapp Number, 24/07 Booking
VIP Call Girls Ankleshwar 7001035870 Whatsapp Number, 24/07 Booking
 
BSides Seattle 2024 - Stopping Ethan Hunt From Taking Your Data.pptx
BSides Seattle 2024 - Stopping Ethan Hunt From Taking Your Data.pptxBSides Seattle 2024 - Stopping Ethan Hunt From Taking Your Data.pptx
BSides Seattle 2024 - Stopping Ethan Hunt From Taking Your Data.pptx
 
(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
(ANJALI) Dange Chowk Call Girls Just Call 7001035870 [ Cash on Delivery ] Pun...
 

Dynamic Dispatch Waves for Same-day Delivery

  • 1. Dynamic  Dispatch  Waves  for   Same-­‐Day  Delivery   Mathias  Klapp,  Alejandro  Toriello,   Alan  Erera   School  of  Industrial  and  Systems  Engineering   Georgia  Tech     UC-­‐Berkeley  ITS  Friday  Seminar   February  20,  2015  
  • 2. What  to  remember   1.  Last-­‐mile  home  delivery  logis=cs  costly  due   to  poor  scale  economies,  and  same  day   delivery  adds  to  challenge   2.   Dynamic  vehicle  dispatch  strategies  for  SDD   systems  may  provide  significant  value  over   fixed  wave  strategies   3.  Simple  rollout  policies  produce  high  quality   dynamic  solu=ons  in  idealized  seJng    
  • 3. This  talk  is  not  about…  
  • 4. Last-­‐mile  home  delivery   •  Weak  scale  economies   – Ton-­‐miles  /  operator-­‐hour  compara=vely  low   – Small  vehicles,  opera=ng  cost  inefficient   Distribution center Consumer delivery locations
  • 9. Pick/pack/load  and  vehicle  dispatch   •  Both  benefit  from  order  batching   – Pick  density  for  warehouse  opera=ons   – Stop  density  for  vehicle  rou=ng  opera=ons  
  • 10. Pick/pack/load  and  vehicle  dispatch   •  Both  benefit  from  order  batching   – Pick  density  for  warehouse  opera=ons   – Stop  density  for  vehicle  rou=ng  opera=ons   Distribution center dense = shorter travel time per delivery
  • 11. Pick/pack/load  and  vehicle  dispatch   •  Both  benefit  from  order  batching   – Pick  density  for  warehouse  opera=ons   – Stop  density  for  vehicle  rou=ng  opera=ons   Distribution center sparse = longer travel time per delivery
  • 12. Next-­‐day  vs.  same-­‐day     yesterday today time orders arrive Next-day Local Distribution System order pick, pack, and load vehicles for delivery dispatched
  • 13. Next-­‐day  vs.  same-­‐day     yesterday today time orders arrive Same-day Local Distribution System order pick, pack, and load vehicles for delivery dispatched
  • 14. Pick/pack/load  batching  economies?   yesterday today time orders arrive Same-day Local Distribution System order pick, pack, and load vehicles for delivery dispatched many orders arrive after first picks must be made
  • 15. Dispatch  batching  economies?   yesterday today time orders arrive Same-day Local Distribution System order pick, pack, and load vehicles for delivery dispatched some vehicles should be dispatched before all orders are ready
  • 16. Vehicle  dispatch  challenges   •  Each  vehicle  dispatched  mul=ple  =mes  during   opera=ng  day  (10-­‐12  opera=ng  hours)   – When  to  dispatch  vehicles?   •  Tradeoffs  between  wai=ng  to  dispatch,   dispatching  long  routes,  dispatching  short   routes   – When  to  wait  to  accumulate  stop  density?   – Which  orders  to  serve  with  each  vehicle   dispatch?  
  • 17. Dynamic  Dispatch  Waves  Problem   •  Determine  dispatch  epochs  dynamically   •  Explore  tradeoffs  ini=ally  with  single  vehicle   and  simplified  geography   time wait dispatch 1 dispatch 2 dispatch 3
  • 18. Simplified  geography:  stops  on  line   di •  Order  loca=on     – round-­‐trip  travel  =me   from  DC   •  No  stop  =me   •  Dispatch     – serves  all  ready  orders   di di {j : dj  di}
  • 19. Order  ready  Mme  process   di time Ready is picked, packed for loading (no duration) T 0⌧i Orders served and vehicle back to DC by time 0
  • 20. Order  ready  Mme  process   di time Ready orders for first dispatch of day T 0⌧i
  • 21. Order  ready  Mme  process   di time Orders that come available later in operating day, and unknown when planning at time T T 0⌧i
  • 22. Dynamic  Dispatch  Waves  Problem   •  Each  =me  vehicle  at  distribu=on  center,  decide:   –  Whether  to  dispatch  vehicle,  or  wait   –  If  dispatched,  which  unserved  ready  orders  to  include   in  the  route   •  Given  set  of  poten.al  orders   –  Round-­‐trip  dispatch  =me   –  Stochas=c  =me  (or  wave)  when  order  ready   –  Penalty  if  order  remains  unserved   •  Operate  to  minimize  total  cost  of  all  dispatches   plus  total  unserved  order  penal=es   N = {1, ..., n} di ⌧i i
  • 23. Dynamic  programming  formulaMon   for  DDWP  on  the  line   •  State:   –  Number  of  remaining  waves,   –  Ready  and  unserved  orders,     –  Poten=al  orders  not  yet  ready,   •  Ac=ons:    wait  one  wave,  or  serve   –  Cost:   –  Possible  dispatch  ac=ons:   –  Must  return  by  0:   •  At  end  horizon,  pay  penal=es  for  unserved  orders   (t, R, P) t R P S ✓ R |R| x = maxi2S di x  t
  • 24. Dynamic  programming  formulaMon   for  DDWP  on  the  line  
  • 26. DeterminisMc  DDWP  on  line   •  Request  ready  =mes  known  in  advance,  but   requests  cannot  be  served  before  ready  =me   •  Proper=es  of  op=mal  solu=on   –  Dispatch  lengths    x    strictly  decreasing   –  No  wai=ng  a]er  first  dispatch    
  • 27. DeterminisMc  DDWP  on  line   •  Request  ready  =mes  known  in  advance,  but   requests  cannot  be  served  before  ready  =me   •  Proper=es  of  op=mal  solu=on   –  Dispatch  lengths    x    strictly  decreasing   –  No  wai=ng  a]er  first  dispatch    
  • 28. DeterminisMc  DDWP  on  line   •  New  DP  state:  remaining  waves  t,  length  d  of  prior  dispatch   •  Recursion   O(n2 T)
  • 29. Using  determinisMc  DDWP   •  Es=ma=ng  an  a  posteriori  cost  lower  bound     – Average  cost  for  sample  of  order  realiza.on  days   – Any  dynamic  policy  for  stochas=c  DDWP  can  have   no  lower  expected  cost   •  Building  a  priori  policy  solu=ons  to  the   stochas=c  DDWP  
  • 30. A  priori  soluMon   •  Before  first  dispatch  (i.e.,  at  wave  T),  find     complete  set  of  vehicle  dispatches:   •  Theorem   {(xk , tk )} Optimal a priori solution is solution to deterministic DDWP where each order i replicated for each wave t 2 {T, ..., 1} with known ready time t and penalty i Pr(⌧i = t)
  • 31. Dynamic  policies   1.  Implement  a  priori  solu=on,  but  adjust  during   opera=ons   –  Shorten,  delay,  and  cancel  some  dispatches   2.  Rollout  using  a  priori  solu=ons   – Execute  first  decision  in  adjusted  a  priori  solu=on   – Build  new  a  priori  plan  any  =me  vehicle  at   distribu=on  center,  using  new  informa=on  
  • 32. Experiment  1   •  Request  ready  =me  process   – Condi=onal  arrival  likelihoods                  each  wave   •  Request  loca=ons  and  penal=es   – Loca=on  discrete  uniform  up  to  a  maximum   – Penal=es  discrete  uniform  on  quarters  of   •   20  random  instances  for  class   – r  measures  =me  flexibility   ✓i T ` ` ✓ n, `, r = T ` ◆
  • 33. Experiment  1:  Results  AvgGaptoaposteriorilowerbound
  • 34. Experiment  1:  Results  AvgGaptoaposteriorilowerbound
  • 35. Experiment  1:  Results,  r  =  2  AvgGaptoaposteriorilowerbound
  • 36. Dynamic  policies  via  ALP   •  Dual  LP  reformula=on  of  Bellman’s  equa=on   – massive  LP:  exponen=al  variables,  constraints   max ERT [CT (RT , N RT )] C0(R, P)  X i2R i Ct(R, P)  EF t 1 [Ct 1(R [ Ft 1, P Ft 1)] Ct(R, P)  d + EF t d [Ct d(Rd [ Ft d, P Ft d)]
  • 37. Dynamic  policies  via  ALP   •  ALP  restric=on  provides  lower  bound,  and   poten=ally  useful  approxima=on  of  C   •  Restrict  C  :   •  “cost  of  unserved  known  requests”,  “cost  of   unserved  poten=al  requests”,  “value  of   remaining  waves”   Ct(R, P) ⇡ X i2R at i + X j2P bt j tX k=1 vk
  • 38. Dynamic  policies  via  ALP   •  Proposi=ons   – Using  this  restric=on  in  dual  LP,  the  ALP  lower   bound  LP  requires                                  variables  and                                       constraints   – (*)  For  determinis=c  problems,  the  ALP  lower   bound  is  =ght,  equal  to  op=mal  cost   •  Hybrid  ALP-­‐A  priori  rollout  policy   – Use  a  priori  rollout  first,  then  switch  to  ALP  rollout   later  in  opera=ng  period   O(nT) O(n2 T)
  • 39. Experiment  1:  Results  AvgGaptoaposteriorilowerbound
  • 40. Experiment  2   •  New  ready  =me  process   – p:  Likelihood  ready  by  T   – q:  Likelihood  of  no  request   – Remaining  likelihood  discrete  uniform:   •  Request  loca=ons  and  penal=es  as  before   •  20  random  instances     (n = 20, ` = 10, r = 3) µi µi + vµi v
  • 41. Experiment  2:  Results  vs.  q  -­‐  p  AvgGaptoaposteriorilowerbound
  • 42. Experiment  2:  Results  vs.  v  AvgGaptoaposteriorilowerbound
  • 43. ObservaMons:  1   •  Dynamic  soluMons  valuable   – Dispatching  scheme  from  A  priori-­‐rollout   approach  usually  provides  significant  savings  over   instance-­‐specific  A  priori  solu=ons   – Schemes  with  fixed  dispatch  waves  could  be  no   befer  in  this  seJng   time wave I wave II wave III
  • 44. ObservaMons:  1   •  Fixed-­‐but-­‐flexible  dispatch  waves  (?)   – Fixed  planning  waves  useful  to  DC  pick/pack/load,   and  for  customer  order  management   – Design  and  performance  of  a  fixed-­‐but-­‐flexible   dispatch  wave  system?   time wave I wave II wave III
  • 45. ObservaMons:  2   •  LocaMons  on  line  creates  maximum  batching   benefit   – Compounded  by  assump=on  of  no  fixed  stop  =me   required  per  delivery   – Incen=ve  to  wait  and  batch  may  be  too  strong   – Inves=ga=ng  problems  with  fixed  stop  =mes  and   two-­‐dimensional  delivery  loca=ons    
  • 46. ObservaMons:  2   T 0⌧i latest wave and dispatch duration
  • 47. Other  extensions   •  MulMple  vehicles  per  delivery  zone   – How  to  coordinate  dispatch  waves  for  two   vehicles  serving  a  single  zone?    Other   configura=ons?   •  Customer  order  management   – Reject/not  offer  same  day  delivery  op=on   dynamically  as  orders  are  received  
  • 48. What  to  remember   1.  Last-­‐mile  home  delivery  logis=cs  costly  due   to  poor  scale  economies,  and  same  day   delivery  adds  to  challenge   2.   Dynamic  vehicle  dispatch  strategies  for  SDD   systems  may  provide  significant  value  over   fixed  wave  strategies   3.  Simple  rollout  policies  produce  high  quality   dynamic  solu=ons  in  idealized  seJng