Kasil Hariharan

1
Team Mission PossibleEnd of Semester
PresentationFall 2003

• Kasil Hariharan
• Junjie Lu
• Jaime Oviedo
• 
  Ming Zhao

Mission Possible
Team Mission Possible EOSP 12/09/2003
2
Outline

• The Team
• The Project
• The Process
• The Experience
• The Future

Mission Possible
Team Mission Possible EOSP 12/09/2003
3
Our Team

• Core team members
• Kasil Hariharan
• Junjie Lu
• Jaime Oviedo
• Ming Zhao
• Client
• William P. Milam from the Ford Motor Company
• Mentors
• Clifford Huff
• Marco Schumacher
• Technical advisor
• Bradley Schmerl

The team
4
Our Project Background

• Model-based design in the automotive industry
• Automated model compilation
• Synergys development
• Automated Model Compiler (AMC)
• Built on top of Acme Studio

The project
5
AMC v1.0
Information flow direction
Source Team Synergys Final Presentation v2.3
Software system
The project
6
Project Background (cont)

• The need for higher levels of component reuse
• The current reuse practices at Ford

The project
7
Component Reuse Example
Manual Transmission Architecture
Automatic Transmission Architecture
Gearbox Frame
Final Drive Frame
Source Fords studio follow-on presentation
The project
8
AMCi Project Objectives

• Integrate AMC with the SynchroMod model
  repository
• Enhance AMC editor, add support for hierarchical
  navigation, units, types and categories
• Address the reported defects in AMC

The project
9
AMCi Project Scope
MatLab/Simulink
Distiller
Brewer
Team Synergys AMC Module Acme
SynchroMod Model Repository
Focus of scope
Information flow direction
Software system
The project
10
AMCi Project Scope

• 16 unfulfilled requirements from Team Synergy AMC
  v1.0 SRS
• 163 new functional requirements
• Emphasis in the quality of the documentation

The project
11
Project Summary

• Synergys AMC tool supports model based design in
  the automotive industry
• AMCi is an enhancement of AMCv1.0, that
  integrates with the SynchroMod tool
• The team will address the reported defects of AMC
  v1.0

The project
12
The Process

• We adopted TSPi as our management process
• We studied XP and RUP as process alternatives
• We incorporated the notion of cycles to organize
  our schedule
• We have launch meetings at the beginning of a
  cycle and postmortem meetings at the end of it
• We assigned TSPi roles to each team member
• We are increasingly incorporating TSPi forms into
  our activities

The process
13
TSPi Cycles

• During the semester we had 3 TSPi cycles
• Cycle one (2 weeks)
• Very ad hoc launch meeting (little preparation)
• Cycle two (2 weeks)
• Very long launch meeting (6 hours / 30 man-hours
  preparing, 2 hours at the launch meeting)
• Cycle three (3 weeks)
• Much less time, more organized and efficient (2
  hours / 6 man-hours preparing, 1 hour at the
  launch meeting)

The process
14
Risk Management -1

• We defined our risk process and we followed it
• We had three risk meetings in the semester
• We defined the cycle objectives based on the main
  risks
• We improved our risk management process through
  the semester

The process
15
Risk Management -2

• The team discusses the current risk list and
  identifies new risks

• The risks are prioritized according to impact,
  probability and timeframe

• The team identifies mitigation strategies and
  determines the actions to take

• The team uses the risk meeting results as an
  guiding input to cycle objectives

The process
16
Our Last Top 5 Risks

• Requirements unstable
• Development tools too complex
• Development based on incorrect requirements
• Incomplete requirements
• Legacy code documentation lacking

The process
17
Quality Management

• We defined our quality process and we strive to
  follow it
• We performed formal reviews for all our main
  deliverables (Average time per review 3.5 hours /
  14 man-hours)
• We performed several informal reviews on all the
  documents we produced

The process
18
Process Summary

• We adopted TSPi as our process model after
  studying different alternatives
• We used what we learned in management class to
  define processes that we are successfully
  followed

The process
19
Our Experience

• The challenges and difficulties in the first half
  of the semester
• The middle of semester resolutions
• The improvements second half of the semester
• The teams achievements

The experience
20
The First Half of the Semester

• Mission Impossible
• Inefficient communication with the client
• We were reactive instead of proactive
• We were course centric - driven by course
  deliverables and deadlines
• We were failing to integrate the part-time member
  
• We were not using the resources available
• We had no idea where we stood

The experience
21
The Middle of the Semester

• We started using TSPi cycles
• We received feedback from the MSOP presentation
• We became project centric

The experience
22
MSOP Feedback

• Use the available resources
• Partition your learning strategy
• Make a categorization of the defects

The experience
23
The Second Half of the Semester -1

• We contacted Synergy team members to help us
  understanding the problem
• Digital motion screen capture of the description
  of the API elements in AMC
• Digital motion screen capture of a demo about the
  operations of AMC

The experience
24
The Second Half of the Semester -2

• We started our efforts to detect and classify the
  bugs
• Two major bugs reported
• One wasnt a bug, but reflected problems in the
  documentation
• The other was located in the export to Simulink
  mechanism of AMC

The experience
25
The Second Half of the Semester -3

• We partitioned our learning strategy
• Each team member will become a specialist in one
  area
• The team will develop a plan of action for each
  specialist next semester

The experience
26
The Second Half of the Semester -4

• Team/client communication improved greatly
• We use Fords BT Web Conference, after exploring
  several alternatives
• We use Hypercam/ACS Capture to record critical
  interview parts digitally (both video and audio)
• We have improved our team website so now is a
  better source of information for the client, the
  mentors and the team members

The experience
27
The Second Half of the Semester -5

• We established weekly Saturday meetings
• We started using TSPi cycles to identify,
  schedule and trace our tasks and progress
• We started using Work Breakdown Structure (WBS)
  to define milestones with entry and exit criteria

The experience
28
Our Achievements

• We have established an efficient means of
  communication with the client
• We now know where we stand, where we are going,
  and what we need to do to get there
• We work as an effective team
• We have moved from near chaos to an improving
  process

The experience
29
Areas of Improvement

• We need to be more proactive
• We need to improve our planning
• We can continue to improve our internal and
  external communications
• We need to improve our projected vs actual effort
  tracking

The future
30
Lessons Learned

• A good process moves you towards success, lack of
  process doesnt let you move, and shelf ware
  moves you away from success
• Communication is critical (we knew this, but now
  we really understand it)
• Quality takes time and effort, and a lot of both
• Its not enough to have good resources, you have
  to use them

The future
31
Future Steps -1

• Perform a fall semester postmortem
• Perform a team-based mentoring for the persons
  assuming new roles
• Conduct a mini-SRE with our client facilitated
  Ray Williams (SEI)
• Continue improving our processes
• Explore the lower level details of AMC and
  SynchroMod

The future
32
Future Steps -2

• Design and architecture AMCi
• Address reported defects
• Work on our areas of improvement
• Refine our SOW and SRS

The future
33
Summary

• We have solved most of the issues that we had in
  the first semester
• We have learned to work harmoniously as a team
• We acted upon the feedback provided in the MOSP
• We know that Quality is Job 1

Mission Possible
Team Mission Possible EOSP 12/09/2003
34
Questions?
Mission Possible
Team Mission Possible EOSP 12/09/2003
35
Our Advantages

• The Synergy team members are always willing to
  help us
• We also have an the documentation of the first
  version
• Our client has worked with studio teams before
  and understand how the MSE Studio works
• We have an excellent level of flexibility in both
  the SRS and the SOW

The experience
36
Roles

• Process/Quality Manager Junjie Lu
• Planning Manager Kasil Hariharan
• Team Lead Jaime Oviedo
• Client Liaison Ming Zhao

The process