Run Optimization Using Simulation Optimization to Develop More

Run Optimization Using Simulation & Optimization to Develop More Efficient Bus Run Optimization 2010

Objectives For This Session • Explain concepts of: Simulation & Optimization • Practice the use of: Simulation & Optimization

NOTE All manual references refer to the: 2009 TIMS/Edulognt User Manual This manual be downloaded by following these directions: WWW. NCBUSSAFETY. ORG TIMS – School Bus Routing Documents

Improvements • One set of tools for manipulating Runs and Routes are used in Simulation and Regular Data. • Optimization functions are limited to the creation of new “starting points” --- all the other tools used for cleanup and refinement are the same as those used elsewhere in the system.

Sim/Opt Simulation and Optimization allow experiments with run and route changes. – These can be all manual changes, or changes generated by optimization. – You can decide whether to confirm to your regular data, save it as a saveset, or discard. – Run and Route Optimization are only available through Simulation.

Entering Simulation and Optimization in Edulog. nt Database – Live, WI, SS, Fall, etc. Edulog. NT Simulation – Select Stops/Runs Optmization-Create New Runs

Exiting Simulation and Optimization in Edulog. nt Optmization Create new Runs, then Save Active Saveset Simulation Modify, then Save Active Saveset Edulog. NT Database – Live, WI, SS, Fall, etc.

Multiple Simulations! • Simulation allows you to work with multiple simulation “savesets” Essentially, you can save any simulation under whatever name you choose and come back to it later to continue working

Simulation in Edulog. nt • Multiple scenarios can exist in Simulation. • Savesets can be confirmed or saved under different names for future use. . • All changes in a simulation may be manual. • Changes you make in simulation only affect the dataset that you are working in. E d u l o g. n t ulation im S m ir f n o C 3 In Saveset Dataset: Live, WI, Fall, etc. Saveset 3 Saveset 2 Saveset 1 Simulation Datasets or ‘Savesets’

Optimization Data in Edulog. nt Simulation • Optimization scenarios can be written to various Simulation Savesets • Optimization becomes just another tool available when creating simulations E d u l o g. n t Post or Save ACTIVE SIM saveset Saveset 3 tion Saveset 2 ula Confirm Sim Saveset 1 Post or Save ACTIVE Opt saveset Dataset: Live, WI, Fall, etc. Saveset 3 Saveset 2 Saveset 1 Optimization

Back Up Your Data! • The number of savesets and locations of SIM and OPT data can be confusing • Back up first in case you confirm something into your regular data by mistake!

Back Up Your Data! Back Up Your Data! Back Up Your Data!

Steps In Run Optimization • Load the data and define the problem • Use Run Opt to create a new set of runs • Evaluate • Make Manual Changes • Confirm the runs back to your real data

Example of How Sim/Opt is Used • You want to “clean up” runs for several schools in the western part of the district by reducing stops, eliminating crossing routes, and possibly reducing the number of buses needed….

Using Sim/Opt • Go into Simulation and load the stops and runs you want to work with. • Make manual changes to clean up runs that aren’t too bad. • Go into Optimization and try to reduce the number of buses used in one school from 6 to 5.

Using Sim/Opt • You think you have the runs in good shape for the school, but you want to go over the solution with someone at the school before you confirm them to your data…. • So, you save the first Opt solution under the school’s name and start a new Opt solution for the next school

Using Sim/Opt • Once you review your Opt solutions you confirm each one to your Simulation • Now, You can review your Simulation and make any additional changes before you confirm the whole thing to your regular data

Using Sim/Opt • Optimization may or may not be used once you are in Simulation Saveset 3 DYN • If you do use OPT, Regular you confirm OPT Data to SIM, then SIM to regular data lation Confirm Simu ta To Regular Da Saveset 2 Saveset 1 “Active” Confirm one Or more OPT Scenarios to Simulation Optimization 1 2 3

Entering Simulation • The yellow “S” the Simulation B takes you into • Only after you enter Simulation can you proceed with Run Optimization

Run Optimization/Simulation Step 1 – Defining the Problem Load the Data in Simulation Define the Problem in Optimization Step 2 – Creating the Initial Runs Step 3 – Evaluating the Runs Step 4 – Making Manual Modifications Step 5 – Confirming the Solution

Entering Optimization • Optimization Steps – Loading Data • When to load data? • You can only work with data loaded properly into your Simulation

Run OPT - Problem Definition The first screen you will see asks you to decide basic questions about your Run OPT session. • AM or PM • Assigned Load or Head Count • Deadhead Mode

Run OPT - Problem Definition • Next, you will determine which of your loaded schools will be included in this problem • Click the Add button to add your selections to the stops to use

Run OPT - Problem Definition • If you are working with more than one school, you would select it from the pull down, and add it as well • The second school would then appear in the right window

Loading Data • Load School Stops! • Run Optimization won’t function until at least one school stop has been loaded (i. e. 304. 000). • If you only load cluster stops, you will have to manually load schools before you can use Opt. • Load Stops • Load the set of stops you want to work with from a worklist, range, or individually by stop number.

Selecting Stops • You must select stops before you can proceed • You can do this by selecting the top stop and scrolling down to the bottom of the window and holding <shift> down while selecting the last stop

Selecting Stops via the Map • Selecting on Map Operations brings up a different “Stops to Use” window • This allows you to pick stops by Point or Polygon with options for Adding or Removing stops • Click Return when you have finished selecting stops

Finishing Problem Definition Note the number of stops and load selected before selecting OK

Run Optimization Step 1 – Defining the Run Optimization Problem Step 2 – Creating the Initial Runs Use Run Opt to Create a New Set of Runs Step 3 – Evaluating the Runs Step 4 – Making Manual Modifications Step 5 – Confirming the Solution

Global Moves Select the number of buses and capacity • Click on Add • You can enter several different sets of buses/capacities before clicking OK • Buses/capacities will appear in the right window

Global Moves • ‘Remove’ option on this page ONLY refers to a run displayed and highlighted in the Right window. Once you click OK, you cannot edit these buses from here, you can only additional buses More about editing, changing, deleting, etc. later. Run Optimization 2008

Runs are now available to work with

Run Optimization Step 1 – Defining the Run Optimization Problem Step 2 – Creating the Initial Runs Step 3 – Evaluating the Runs Step 4 – Making Manual Modifications Step 5 – Confirming the Solution

Checking Runs • Runs > Open Run Summary shows you stats about one run at a time • Check each run by selecting the mouse and clicking on the run on the map, or select the run ID from the pull down menu

Checking Runs Bet ya couldn’t find it!!! Gotta Look Closely!!!

Improving Runs • Use the Improve Runs function under Global Moves until no moves are made. • Runs shown have been through 3 sets of the improve function.

Run Optimization Step 1 – Defining the Run Optimization Problem Step 2 – Creating the Initial Runs Step 3 – Evaluating the Runs Step 4 – Making Manual Modifications Step 5 – Confirming the Solution

Runs Changing/editing the number of buses and capacity You may ADD additional buses by returning to GLOBAL MOVES>Run Generation Note that the runs you previously entered are STILL active. The ones that you select now will be ADDED to these.

Runs Changing/editing the number of buses and capacity • You may delete buses(runs) in your senario by going to RUNS>Delete • It will list all of the Current runs, highlight the ones that you want to delete or click on the mouse and point to the runs to delete and click on ‘Delete’ Note: You do NOT lose the stops, they are put ‘back-in-play’

Runs Changing/editing the number of buses and capacity • You may edit the existing runs in your senario by going to RUNS>Cap. Time • It will list all of the Current runs, highlight one and select ‘Edit’

Runs Changing/editing the number of buses and capacity RUNS>Purge Will delete any runs that have no stops. If you over-estimate the number of buses needed, this will take away the extras. Run Optimization 2008

Making Edits to Your RUNOPT Runs Stops > Deassign allows you to remove several stops at once.

Inserting Stops On New Runs • Inserting stops within run OPT is not very refined at this time • Once you insert, you can run “improve runs” to make better sequences

Making Edits • Tools are limited in Run OPT for these kinds of edits, however, the basic assign and deassign options in OPT are great for doing basic run cleanup Note: Run Opt only allows individual stop assign/deassign. Use SIMULATION to have the Polygon Feature available.

Saving OPT Runs to Simulation • Saving runs back to simulation will allow you to make edits easily • Once you do this, you must confirm your changes from the first session of run OPT to a file or your data before you can re-enter run OPT

Run Optimization Step 1 – Defining the Run Optimization Problem Step 2 – Creating the Initial Runs Step 3 – Evaluating the Runs Step 4 – Making Manual Modifications Step 5 – Confirming the Solution Confirming to Your Real Data

Make Sure • Once you have confirmed your runs from Simulation back into your real data, YOUR REAL DATA IS CHANGED! • You can only go back if you have a good BACKUP!

Recap How You Should Use Run OPT • You should think of OPT as a tool to be used as part of simulation • It will give you a starting point for a particular set of runs • Simulation allows you to continue to work with the runs to clean them up and make other needed edits

Recap What You Do Where • Optimization – Basic run generation, deassigning and reassigning stops • Simulation – Local moves with improved run data (no actual run directions, but time estimates in easier to use format) • Regular Data – Real run directions and edits to run directions and reports!

Get Started • Enter Simulation by pressing the yellow button in the tool bar • Load a Range of stops for School 310, from 310. 000 to 310. 999 • Now enter Run OPT by going to File > Enter Optimization > Run OPT Practice

Setting the Problem • See Pages 386 -391 in the Run Optimization Guide for set by step instructions on the use of the software • Go ahead and follow those steps for the school 310 stops mentioned • Stop at step 35, Generating the Solution Run OPT Practice

Run OPT Practice Generating a Solution • The documentation goes right into how to generate runs • It does not give any explanation of how you decide on the number of buses you need. • How do you know how many bus runs to create? Run OPT Practice

How Many Buses? • Run the report under Reports > Generate Solution Reports • This will tell you how many stops you have chosen, and how many students are assigned to them. • Look in your c: users%username%tmpRUNOPT. RPT NOTE: Newer systems may have the ‘USERS’ folder on the D: drive Run OPT Practice

How Many Buses? • Suppose you find the following: # of stops = 111 Run Load = 241 • Divide the Run Load by the Bus Capacity: 241/54=4. 46 • Let’s say that means you want to Run 5 buses…. Run OPT Practice

How Many Buses? • Now you can proceed with the steps in the documentation under Generating a Solution • Note that it is better to adjust the bus capacity maximum so that it is less than or equals the total students to assign, so that all the runs are balanced Run OPT Practice 5 x 54 = 270 Since you had only 251 students, adjust the maximum to 50 so the stops will spread evenly among the runs

Generate Runs • Continue with the documentation steps to create runs • Stop at step 47 (Pg. 394) Run OPT Practice

Generate Runs Run OPT Practice

Improving the Solution • Hide layers as instructed in step 48 on page 394. • Use the Global Moves > Auto Improve function • Now you have a preliminary set of runs Run OPT Practice

Saving the Run OPT Problem • Follow the documentation instructions to name your solution set (pg 395) • Call your solution set “ 310 first attempt” • It is a good idea to have a saved solution in case you run into problems and want to go back to your starting point Run OPT Practice

Making Manual Modification • The next slides show an area where there runs need to be cleaned up • First the stops are de-assigned from one run, then assigned to another Run OPT Practice

Making Manual Deassign the Red. Modification Run Stops Run OPT Practice

Assign them to the Blue Run OPT Practice

This Speeds Up the Cleanup Run OPT Practice

Practice Cleanup • Look at your own solution • Find an area where stops need to be moved from one run to another – Use • Stops > De-assign, then • Stops > Assign to move those stops from one run to another Run OPT Practice

Practice Cleanup • Use the Run OPT Stats to help you see what you are doing to the load of your runs as you make changes • Run OPT stats is minimized in the lower left of the Edulog. nt window when you generate runs • You can show or hide it as needed as you work with the runs Run OPT Practice

Practice Cleanup • Now review the rest of your solution • Find other areas where you want to move stops • Check your progress by going to: – Runs > Open Run Summary or – by generating new reports for runs under Reports > Generate Solution Report Run OPT Practice

Saving the Run OPT Problem - Again • Follow the documentation instructions to name your solution set (pg 395 steps 52 -55) • Call your solution set “ 310 final cleanup” • Now you have two savesets… the first pass from OPT and the work you have done cleaning things up Run OPT Practice

Save the Run OPT Solution to Simulation • Skip ahead to pg. 397 and follow steps 67 -76 • This will put you back in Simulation Mode • Remember, you go from regular data to Simulation to OPT • To leave OPT you go to Simulation Run OPT Practice

Making Manual Modification • You are now in Simulation Mode, with the Run OPT runs available to be edited • You can continue to do more edits to the OPT solution here, if you want Run OPT Practice

Simulation to Real Data Making Manual Modification • Until you confirm your data, your Simulation is just an idea. • For instructions on how to confirm your changes, go to pg. 399 of your documentation. • For today, we will just keep it an idea… Run OPT Practice

Run OPT Practice 2 • School 312 is a county-wide HS • If you follow the documentation, you should find 374 stops with 626 students assigned • Currently, the district runs about 21 AM runs with a load of 30 students • Create 19 Runs with a load of 33 students each Run OPT Practice 2

Run OPT Practice 2 Hints • Double-check the numbers to make sure your data matches the example. • If it doesn’t adjust your solution to match your situation. • Go ahead and save a preliminary solution, then save periodically to new names as you make progress – that way, you can go back to various stages and try different strategies Run OPT Practice 2

Run OPT Practice 2 Hints • Practice checking the status of your manual changes by going to Runs > Open Run Summary, scrolling over to the run load columns. • Periodically check the overall status of runs by going to Reports > Generate Solution Reports. Run OPT Practice 2

Edulog. nt Run Optimization Using Simulation to Develop More Efficient Bus Runs