Warehouse and Distribution Science
CHAPTERS 10 & 11 ROUTING TO REDUCE TRAVEL AND WORKFLOW AND BALANCE
Routing
CHAPTER 10
Learning Objectives
Understand the complexities of implementing pick
routing Explain why straight line and serpentine paths are usually good Understand the relationship between product placement and travel time in picking & some rules of thumb about best placement of popular skus Understand why developing optimal pick routings is a difficult problem to solve and not included in most WMS Implement by hand a simplified version of Ratliff and Rosenthals algorithm for a small pick problem
Shortest Route Problem
What is the shortest
route or path from a given node in a network to all other nodes?
Mapquest Deliveries Warehouse picking Critical path in project management More
Traveling Salesman Problem
http://www.tsp.gatech.edu/games/tspOnePlayer.html
Complexity of the TSP
In the theory of computational complexity, the
decision version of the TSP belongs to the class of NP-complete problems. Thus, it is likely that the worst case running time for any algorithm for the TSP increases exponentially with the number of cities.
TSP at UCF for ILL
UCF Main campus Objectiveobtain route
with minimum travel distance to visit all departments to deliver inter-library load (ILL) material Start and End at Library Nodes - Department buildings (X, Y) locations Assume One Vehicle
TSP Model
Objective: To minimize the total travel distance of the
route
Minimize
dij * Xij
i 1 j 1
Constraints:
1.
Vehicle should leave each node (including library)
Xij = 1
i 1
2.
Vehicle should visit each node and return to the depot (library)
Xij = 1
j 1
3.
Subtour elimination: Vehicle should make only one complete tour. Xij |S| - 1 S N
iS jS
Results of the Traveling Salesman Tour
HPA1 CAS HPA2
Communications BHC 75 Colbourn Hall 95 18 Education 21 14 2 12 5 54 45
87
Engg 1 & 2
80
90
Creol
91 94 53 BA 2
20
BA 1 HP Hall Bio. Sci. Library
Math & Physics
Chemistry
CSB
What if have two vehicles?
2. Two routes instead of one Cluster first, route second
CLUSTER 1
CLUSTER 2
Two Optimal Routes w Capacity Constraint
ROUTE 2
ROUTE 1
75
87
80
95 18 21 14 2 12 5 54 45
90
91 94 20
53
So what about TSP for a warehouse?
Rectilinear Aisles Need to know distance
between each pair of locations Not supported by WMS What if the picker doesnt follow suggested route? Think iPad
Heuristics
Global path imposes a sequence that will be
respected by all travel
Known as Probabilistic Traveling Salesman Problem PTSP
Want a pick path that is short (efficient) and simple
to understand Serpentine pick path Branch and Pick
Serpentine Example
Popular skus
Wider aisles
Branch and Pick Example
Branch and Pick Alternatives
Branch and Pick Alternatives
Routing Heuristic
Simplified version of optimal-finding algorithm by
Ratliff and Rosenthal Use Dynamic Programming Assumptions:
Each aisle can be visited only once; To get to the next aisle, the picker can travers the entire aisle or retreat back to the end where they entered;
Example Problem for Routing Heuristic
Shortest Path Problem with 9 Nodes
1st Sub-problem: Aisle 1 to 2
2nd Sub-problem: Aisle 2 to 3
2nd Sub-problem: Aisle 2 to 3
26
16
16
20
2nd Sub-problem: Aisle 2 to 3
3rd Sub-problem: Aisle 3 to 4
3rd Sub-problem: Aisle 3 to 4
4th Sub-problem: Aisle 4 to End
Shortest Path
Another Problem
For any problem, the
cross arcs will be the same The end arcs will change depending on the locations to visit
Can you find the shortest path for this?
How much is optimization worth?
Do we need to optimize routes when there are very
few items (1-3) in a pick order?
Why?
Do we need to optimize routes when there are very
many items (i.e. you need to visit almost every aisle multiple times) in a pick order?
Why?
Where is Optimal Routing most Beneficial?
Warehouses with many items, which are slow
moving Warehouses with orders of moderate size Examples
Hardware distribution centers Building supply warehouses Aftermarket auto parts to dealerships
Workflow & Balance
CHAPTER 11
Learning Objectives
Identify the steps to implementing a bucket brigade
to balance workload Identify the advantages of bucket brigades over other workload balancing policies
Self Organized Teams
Requires no central planning or higher authority.
It is adaptive, i.e. spontaneously adjusts to changes
in the environment
Bucket Brigades in Action
Sample Flow Line
Slower worker
Faster worker
Figure 11.3
Line self-balances at f(x).
Eventual Partition of Work Content
Slowest worker is green; fastest worker is red.
Fastest worker does more of the work.
Some Advantages of Bucket Brigades
Pure pull system, so WIP is controlled
Does not require accurate time studies
Support teams and grouping work cells Simple and easy for each worker to know what to do
Workers can usually put themselves in order of
slowest to fastest
Replacing Zones with Bucket Brigades
Distribution of Average Pick Rate
Before After Bucket Brigades
Bucket Brigades are most appropriate when
All work is based on a single skill, e.g. sewing,
making sandwiches Workers can easily move among stations and take over work in process Demand for product varies significantly
