Intelligent scheduling at NASA Aurora-KSC

Intelligent scheduling at NASA Aurora-KSC

REDUCED PROJECT DURATION VIA INTELLIGENT SCHEDULING FOR SUBMARINE PRODUCTION & MAINTENANCE Pacific International Maritime Conference 2019 Jordan Ciricillo & Jorge Payne General Dynamics Electric Boat

Rob Richards, Ph.D. Stottler Henke Associates, Inc. 1 LESSONS Efficient scheduling is incredibly difficult

Significant scheduling Learn from the human schedulers Model to level of detail required

Reveal & explain the schedule Easy & fast to perform scenarios / what-ifs productivity lost due to

inferior LESSONS Efficient scheduling is incredibly difficult Significant scheduling Learn from the human schedulers

Model to level of detail required Reveal & explain the schedule Easy & fast to perform scenarios / what-ifs productivity

lost due to inferior EFFICIENT SCHEDULING DIFFICULT Model to level of detail required

Capture human reasoning Different scheduling applications generally require different algorithms WHY CANT YOU SEARCH FOR THE BEST ORDER? Ordering options scale as N! Two tasks that can occur in either order (one at a

time) results in two options Three tasks that can occur in any order (one at a time) results in six options Four tasks that can occur in any order (one at a time) results in twenty-four options WHY ORDER MATTERS?

The example below involves jobs using two resources, green and gray B A E C D Schedule 1: B before C

A B C D Schedule 2: C before B A C B

D E E LESSONS Efficient scheduling is incredibly difficult Significant scheduling

Learn from the human schedulers Model to level of detail required Reveal & explain the schedule

Easy & fast to perform scenarios / what-ifs productivity lost due to inferior SIGNIFICANT PRODUCTIVITY LOST DUE TO INFERIOR SCHEDULING

Capture human reasoning Different scheduling domains generally require different algorithms Model to level of detail required SCHEDULING RESULTS REFINERY PROJECT

MICROSOFT PROJECT RESULTS VS INTELLIGENT SCHEDULING RESULTS LESSONS Efficient scheduling is incredibly difficult Significant scheduling

Learn from the human schedulers Model to level of detail required Reveal & explain the schedule Easy & fast to perform scenarios / what-ifs

productivity lost due to inferior LEARN FROM HUMAN SCHEDULERS

When decisions / tradeoffs need to be made, use the expertise of expert schedulers So that the scheduling system reacts as a human expert wants it to E.g., when to work overtime, when to outsource Different scheduling applications generally require

different algorithms Human experts help drive the best decision per algorithms AURORA ARCHITECTURE Tasks, Resources, Constraints Preprocessor Schedule

Objects Task Queue Task Ordering: Pick a Task Process Task Criteria & Instructions Human Directives

Select/Assign Resources Task & Time Window Methods & Knowledge Handle Problems Guidance Knowledge Editors

Fix Problems & Optimize Schedule Problem Resolution Knowledge 16 Propagate Constraints HUMAN KNOWLEDGE DRIVES SETTINGS

1 LESSONS Efficient scheduling is incredibly difficult Significant scheduling

Learn from the human schedulers Model to level of detail required Reveal & explain the schedule Easy & fast to perform scenarios / what-ifs productivity

lost due to inferior MODEL TO LEVEL OF DETAIL REQUIRED Technical constraints (E.g., F-S, F-F, S-F, lags) Resources Labor: Occupation, skills, certifications Equipment, Tools (e.g., cranes)

Usage constraints e.g., tool can only be used for so many hours continuously &/or during a day. Spatial / physical space e.g., job requires a certain location or type of space two elements should (or should not) be next to each other Equipment substitutions equipment down, know & use substitutes CONCURRENT CONSTRAINTS NON-CONCURRENT CONSTRAINT

NON-CONCURRENT CONSTRAINTS: RED ARROWS OUTSOURCE AUTOMATION 2 REMAINING DURATION OVERRIDE 2 LESSONS

Efficient scheduling is incredibly difficult Significant scheduling Learn from the human schedulers

Model to level of detail required Reveal & explain the schedule Easy & fast to perform scenarios / what-ifs productivity lost due

to inferior REVEAL AND EXPLAIN THE SCHEDULE PROJECT NETWORK GANTT CHART COLOR-CODED PER USER DESIRED CRITERIA TEAM ASSIGNMENT DISPLAY

2 NETWORK DIAGRAM SHOWING SINGLE-ELEMENT VIEW OPTION SPLIT VIEW SHOWING GANTT CHART SAME TIME SLICE AS HISTOGRAM, SHOWING ACTIVITIES CONSTITUTING RESOURCE NEED FOR ONE TIME INSTANCE VEHICLE TESTING SCHEDULE WITH BUILD PITCH 3

PHARMA MANUFACTURING MACHINE UTILIZATION REPORT Shows overall utilization of products by line Allows planners to see overall allocation and line balancing AUTOMATICALLY GENERATED EXPLANATION

LESSONS Efficient scheduling is incredibly difficult Significant scheduling Learn from the human schedulers

Model to level of detail required Reveal & explain the schedule Easy & fast to perform scenarios / what-ifs productivity lost

due to inferior EASY & FAST TO PERFORM SCENARIOS / WHAT-IFS Wall clock minimization to run a scenario critical so schedulers will actually run them

Provide ways to graphically compare results WHAT-IF: SAME DEMAND 3 VS 2 LINES 3 WHAT-IF: DEMAND INCREASE 3 WHAT-IF: WORK WEEKEND TO

GET BACK ON SCHEDULE 3 LESSONS Efficient scheduling is incredibly difficult Significant scheduling

Learn from the human schedulers Model to level of detail required Reveal & explain the schedule Easy & fast to perform scenarios / what-ifs

productivity lost due to inferior

Recently Viewed Presentations

  • Introducing Economics and Business

    Introducing Economics and Business

    Intentions. This webinar today is talking about teaching and learning key concepts to introduce Economics to upper Primary students. Introducing economic concepts including resources, needs, wants, production, goods, services, incentives, the basic economic problem of scarcity, opportunity cost, economic decision...
  • John Wayne Gacy - Seneca Valley School District

    John Wayne Gacy - Seneca Valley School District

    John Wayne Gacy Jr by Surjan Stevens. His father was a drinkerAnd his mother cried in bedFolding John Wayne's t-shirtsWhen the swingset hit his headThe neighbors they adored himFor his humor and his conversationLook underneath the house thereFind the few...
  • Chapter 1 Periodic System of The Elements and Atomic Structure

    Chapter 1 Periodic System of The Elements and Atomic Structure

    Arial Verdana Times New Roman Wingdings Arial Black Comic Sans MS Profile CHAPTER 1 PERIODIC SYSTEM OF THE ELEMENTS AND ATOMIC STRUCTURE Teaching Learning Objectives A. Model and Theory of Atom The Greek Model ATOMIC THEORY OF DALTON 2. Dalton's...
  • In-Class Writing Activities

    In-Class Writing Activities

    Two-voice poetry is written for two people to perform. Two-voice poetry is usually written in two columns. Sounds like a dialogue. ... "One day I will find the right words, and they will be simple"—Jack Kerouac, The Dharma Bums. Quote...
  • Invoices - Welcome | SOU Home

    Invoices - Welcome | SOU Home

    Reminders. Before entering the invoice for paying the vendor: Vendor must be in the Banner FIS system BEFORE the invoice can be created. Choose the Vendor's Address carefully . VP address should match what is published on the vendor's invoice...
  • Transformers - sk4education.com

    Transformers - sk4education.com

    Transformers Transformer Voltage Regulation and Efficiency Electrical Machines The output voltage of a transformer varies with the load even if the input voltage remains constant. This is because a real transformer has series impedance within it.
  • Inguru Eta Prozesu Sedimentarioak Harri Sedimentarioak

    Inguru Eta Prozesu Sedimentarioak Harri Sedimentarioak

    Heparinas de bajo peso molecular en Atención Primaria Vol 20, nº 9 - año 2012 Sumario Introducción Profilaxis de la Enfermedad Tromboembólica Venosa (ETV) Indicaciones controvertidas Traumatismos de miembros inferiores Viajes de larga duración Uso en situaciones especiales Ancianos Obesos...
  • Water Quality Test Lab - Denton ISD

    Water Quality Test Lab - Denton ISD

    Water Quality Test Lab. Testing Dissolved Oxygen, pH, Nitrate, Nitrite, Ammonia, and Coliform Bacteria. Purpose. This lab will enhance the understanding of typical water quality parameters in different types of water samples.