The Multidimensional Specimen Measuring System and the Software Engineering Process

1
TheMultidimensional Specimen Measuring
Systemand the Software Engineering Process
Chris Janneck (CJ) Lehigh University M.S.
Comp. Eng. 01 M.A. Theatre 01
2
Project Overview

• Objective
• To create an enhancement for image-capturing
  software to determine the length, size, and other
  secondary features of microscopic structures
  present in the image.
• Originated as a class project for ECE 216
  Software Engineering
• This class project is designed for the students
  to attain real-world Software Engineering
  experience, including finding and working with a
  client

3
Requirements Phase

• Goals
• Analyze clients present situation
• Determine what client needs, not wants
• Interview
• Formal or casual, prepared or ad-hoc
• We had a semi-formal interview with Prof. Swann,
  with many basic questions prepared
• Rapid Prototype
• Designed for better understanding between
  developer and client, not functionality
• We designed an HTML slideshow-like prototype of
  the project

4
Screenshots from R.P.
5
Specifications Phase

• Goals
• To be clear and intelligible to the client
• To be complete, detailed, and well
  inter-referenced
• Contains specific restraints that project must
  follow
• Portability, reliability, accuracy, etc
• The contract between developer and client

6
Our Spec. Doc
Once written, we presented our Specifications
Document to Prof. Swann via our teams website.
This allowed for easy access from any computer,
and also removed the cross-platform barrier
between the developing team and Prof. Swann (I.e.
we own PCs, she uses a Macintosh).
Insert picture of Spec. Doc here
7
Design Phase

• Goals
• Transition from what project does to how it will
  accomplish the functionality
• Action-oriented vs. Data-oriented
• Action Data-flow analysis
• Data Design data structure first, then
  procedures to accommodate data format
• Our Design Document had also been posted on the
  Internet once completed

8
Specifications Snapshots
The MSMS UI Architectural Design
Data Flow chart For the MSMS
9
Implementation Phase

• Goal
• Transition from Design parameters to code
• Choosing of programming language
• Predominant decision due to cross-platform issue
• The microscope interfaces with a Macintosh, and
  very few non-PC developing kits are available on
  campus
• Java is chosen due to high cross-platform
  portability
• Java programs run on a Virtual Machine, not
  directly on the Operating System (Windows98/NT,
  MacOS). Therefore, any computer with this Virtual
  Machine (downloadable from the Internet) can run
  any Java program.

10
Java this aint no sippin coffee

• Java is a highly portable, object-oriented
  language
• Gives rise to its increasing popularity
• Has roots in C
• Basic structure and logic are familiar from
  previous classes
• Object-oriented high-level learning
  classes instead of using techniques
• Most of the summer was spent learning the
  plethora of available classes and functions
  instead of using the programming techniques
  already acquired

11
How Java Works
12
Maintenance Phase

• Goals
• To fix arising bugs, implement new functionality,
  and keep software serviceable for end users
• Begins once final product is presented and
  remains in effect until software is retired
• Unfortunately, this is rarely seen through by
  software developers
• Just as unfortunate, this project never reached
  this phase due to the unexpected wealth of time
  devoted to learning Java

13
Conclusions/Observations

• The initial documentation phases are critical
  in maintaining communication and a common goal
  between developer and client
• More time/energy/expenses in initial phases
  Less time/energy/expenses in final phases
• Orders of magnitude easier to change a few lines
  of text in a Specifications Document than to
  redesign/retest/debug code and re-present a new
  CD to each and every license holder
• Never underestimate the learning curve of new
  software and/or languages

14
A Special Thanks To

• Prof. T. Boult, EECS Department
• Professor of ECE 216 Spring 00
• My ECE 216 fellow team members
• Brian Gilman 01
• David Hepkin 01
• Joel Williams 01
• Prof. Swann, Neuroscience Department
• Our client and contact in the project
• The Howard Hughes Foundation
• For allowing continuing work on this project over
  and above the regular school year

15
References

• ECE 216 Team Homepage
• http//www.eecs.lehigh.edu/jtw6/small-project-tea
  m.html
• Requirements Document
• http//www.eecs.lehigh.edu/jtw6/docs/req.html
• Rapid Prototype
• http//www.eecs.lehigh.edu/cdj2/RPDemo.html
• Specifications Document
• http//www.eecs.lehigh.edu/dahd/spec/
• Design Document
• http//www.eecs.lehigh.edu/dahd/design/