automated course rosteringww1.jeppesen.com/.../crew-and-fleet/presentations/e25_final_sia.pdf ·...
TRANSCRIPT
AUTOMATED COURSE ROSTERING
2
Fleet (as of 01 August 2013)
– 101 aircraft
– Average age of 6 years 9 months
Network
– More than 60 destinations in over 30 countries
Crew strength
– 7800 Cabin Crew
– 2500 Tech Crew
BACKGROUND
3
Background & Objectives
The Enhancement in a Nutshell
– Preparation
– How it works
Progress and Results
Working with the Jeppesen Team
Limitations and Future Steps
AGENDA
BACKGROUND
Existing System with
Good Potential
SIA’s Training
Requirements
Why implement Automated Course Rostering?
• By seniority
• By rank
• By qualification
• Etc.
1. Many training courses
2. Some with complex criteria
Examples:
1. Capable Optimiser
2. Flexible business logic
5
OBJECTIVES
Model courses according to their structure 1
Identify and roster relevant courses to crew
Based on a set of pre-defined criteria for each course,
and working around existing pre-assignments
2
Perform trainer rostering 3
We wanted a model that allows us to
6
PREPARATION
What was needed to achieve Automated Course Rostering?
Standard Crew Information
Crew Training History
Courses available for rostering each
month
Criteria for rostering these courses
Assign new tasks
Modify current manual processes
all necessary data had to
be retrievable within JCR
workflows had to be refined DATA WORKFLOWS
7
HOW IT WORKS
1 2 3 4
Course modelling
& criteria setup
Create
course trips
Sync learning
history
Roster
Optimisation
8
Reduction of manual effort
Reduction in time required for course rostering
~ 2.5% reduction in no. of man-days in a roster cycle
Only need one staff to handle what we needed a
team of 7 to do
New reporting and error-checking functionality
PROGRESS and EXPECTED RESULTS
Where we are now
• Course assignments as a pre-run
• Implemented trainer rostering
• Trainee rostering only for limited courses
• Process of fine-tuning
Where we want to be
• Course assignments to run in with flights
• Most, if not all courses (trainers and
trainees) to be rostered automatically
with minimal manual intervention
1
2
Estimated Benefits from Full Implementation
Progress
9
WORKING WITH THE JEPPESEN TEAM
10
Challenges
• Geographical barriers
• Novelty of project
A Good Learning Experience
• Professional, quality work from the Jeppesen Montreal team, given the
challenging circumstances
• Requirements were understood well and required minimal clarification
• Prompt support for handling issues
Areas for Improvement
• On-site presence would help greatly
• More pre-development discussion/clarification
WORKING WITH THE JEPPESEN TEAM
11
Not all courses can be
modelled with complete
accuracy
Only basic course rostering
ruleset available presently
LIMITATIONS AND FUTURE STEPS
Expand criteria and rostering rule-sets
In order to handle a wider variety of
rostering scenarios, course assignment
criteria will have to be expanded
1
Ongoing process of fine-tuning for
improved rostering efficiency
A work in progress, based on lessons
learned from Production
Limitations Future Steps
2
12
QUESTIONS