1. What is Systems Analysis and Design?

Definition (SAD)

Systems analysis and design (SAD) is the process of:

  1. Examining a business problem
  2. Determining what system will address the problem
  3. Designing that system
PartFocus
Systems AnalysisStudy the current system, identify problems / inefficiencies
Systems DesignCreate a solution: design a system that meets the business requirements

Goal: build effective and efficient systems that solve business problems.

Example — the house analogy (slide 4). Building software is like building a house:

flowchart LR
    A["Family's idea of a house<br/>(business need)"] <--> B["Architect's blueprints<br/>(analysis & design)"]
    B <--> C["Construction site<br/>(implementation)"]
    C <--> D["Finished house<br/>(delivered system)"]

Why a formal process?

  • Failures occur too often.
  • Creating systems is not intuitive.
  • Projects are late, over budget, or delivered with fewer features than planned.

2. The Systems Analyst

Definition (Systems Analyst)

A person responsible for analyzing and designing systems to meet organizational needs.

Key responsibilities

  • Problem-solving — find problems in the existing system or ways to improve the organization.
  • Communication — bridge between users, developers and management.
  • Documentation — clear specifications and requirements for the new system.
  • Facilitating change — smooth implementation and adoption.

⚠ The main goal is not a perfect system — it is to create value for the organization (usually: increase profits). Many systems failed and were abandoned because analysts built a “perfect” system without understanding how it fits the organization’s goals, current business processes and other information systems.

Skills

SkillMeaning
TechnicalUnderstand relevant technologies and design techniques
BusinessUnderstand and align with business objectives
AnalyticalBreak down complex systems
CommunicationTalk to technical and non-technical audiences
InterpersonalLeadership and management
EthicsDeal fairly, protect confidential information

Roles

RoleFocusExample
Systems AnalystMakes sure the IS meets requirements and works correctlyIdentifies what features a new application needs
Business AnalystUnderstands business problems, improves business processesFinds ways to reduce customer waiting time
Infrastructure AnalystTechnical infrastructure (servers, networks, hardware)Ensures servers can support the new application
Change Management AnalystHelps employees adapt to new systemsTrains employees on new software
Project ManagerFinish on time and within budgetAssigns tasks, monitors deadlines, controls costs

3. Systems Development Life Cycle (SDLC)

Definition (SDLC process)

  • Consists of four phases: Planning → Analysis → Design → Implementation.
  • Each phase = a series of steps; each phase produces documented deliverables.
  • Phases can run sequentially, incrementally, iteratively, or in another pattern.
  • SDLC is gradual refinement: each phase uses the previous phase’s deliverables as input and adds detail.
flowchart LR
    P[Planning<br/>WHY build it?] --> A[Analysis<br/>WHO / WHAT / WHERE / WHEN?]
    A --> D[Design<br/>HOW to build it?]
    D --> I[Implementation<br/>Build, install, support]
    I -.-> P

3.1 Planning Phase

Questions: Why build this system? What value does it provide? How long will it take?

StepActivities
1. Project initiationDevelop/receive a system request; conduct a feasibility analysis
2. Project managementDevelop the work plan; staff the project; monitor & control

3.2 Analysis Phase

Questions: Who will use it? What should it do? Where and when will it be used?

  1. Develop an analysis strategy — model the current (as-is) system, formulate the new (to-be) system.
  2. Gather requirements — develop a system concept; create a business model of business data and business processes.
  3. Develop a system proposal — initial deliverable describing which business requirements the system must meet.

3.3 Design Phase

Question: How should we build it?

  1. Design strategy (in-house, outsource, …)
  2. Architecture design (hardware, software, network) + interface design (UI, reports, forms)
  3. Database and file specifications
  4. Program design — what programs to write and what each one does
  • Architecture design + interface design + database/file specs + program design = system specification, handed to the programmers.
  • At the end of design, feasibility and project plan are re-examined; the sponsor/approval committee decides to continue or terminate.

3.4 Implementation Phase

  1. Construct — write code and test it (bugs are very costly → testing is one of the most critical steps).
  2. Install — train the users.
  3. Support — maintenance.

⚠ Don’t confuse the phase questions: Planning = why, Analysis = who/what/where/when, Design = how.


4. SDLC Methodologies

Definition (Methodology)

A formalized approach to implementing the SDLC: it defines the steps to follow and the deliverables of each phase.

Categories by focus

CategoryCore of the system concept
Process-orientedProcess models
Data-centeredData models
Object-orientedBalances both — process + data in one model

Classes of methodologies

flowchart TD
    M[Methodologies] --> S[Structured Design]
    M --> R[Rapid Application Development]
    M --> AG[Agile Development]
    S --> W[Waterfall]
    S --> PA[Parallel]
    R --> PH[Phased]
    R --> PR[Prototyping]
    R --> TP[Throwaway Prototyping]
    AG --> SC[SCRUM]

4.1 Structured Design

Waterfall

Phases strictly in sequence: Planning → Analysis → Design → Implementation. Deliverables are long (often hundreds of pages) and must be approved by the sponsor before moving on.

AdvantagesDisadvantages
Requirements defined earlyDesign must be fully specified before programming
Minimal changes laterVery rigid; long time before delivery
Costly changes if requirements are missed
Poor at handling a changing business environment

Parallel Development

General design first → system split into subprojects designed and implemented in parallel → integrated at the end.

AdvantageDisadvantages
Faster delivery than waterfallDependencies between subprojects (a design decision in one affects another)
Integration can be difficult

4.2 Rapid Application Development (RAD)

Goal: quickly deliver parts of the system to get early user feedback.

  • Strengths: faster development, higher user involvement, better requirement refinement.
  • Risks: user expectations can grow uncontrollably; less detailed documentation.
MethodologyHow it works+−
PhasedAnalysis identifies overall requirements; system delivered in sequential versions — core features first, then improvementsFirst version delivered quickly; users see progress earlyUsers initially work with an incomplete system
PrototypingAnalysis, design, implementation done concurrently and repeatedly; prototype shown to users and improved from feedbackUsers clarify needs quicklyMany changes → messy design over time
Throwaway PrototypingQuick prototype built only to gather feedback/understand requirements, then discarded; real system built from the insightsReduces risk by clarifying requirementsLonger delivery than evolutionary prototyping

⚠ Prototyping evolves the prototype into the final system; throwaway prototyping throws it away.

4.3 Agile Development

Short development cycles, flexibility, frequent delivery of working software, continuous customer–developer collaboration.

  • Core values: working software, collaboration, responding to change.
  • Strengths: adapts quickly to changing requirements; continuous improvement.
  • Criticisms: hard to scale to large systems; limited documentation.

SCRUM

  • Work done in 30-day sprints, with progress reviewed after each.
  • Self-organized teams (they make decisions themselves).
  • Strengths: fast feedback, high transparency, flexible planning.
  • Challenges: scaling to large systems; requires disciplined teams.

4.4 Which Methodology to Use?

Ability to develop systems…WaterfallParallelPhasedPrototypingThrowawayXPSCRUM
With unclear user requirementsPoorPoorGoodExcellentExcellentExcellentExcellent
With unfamiliar technologyPoorPoorGoodPoorExcellentGoodGood
That are complexGoodGoodGoodPoorExcellentGoodGood
That are reliableGoodGoodGoodPoorExcellentExcellentExcellent
With a short time schedulePoorGoodExcellentExcellentGoodExcellentExcellent
With schedule visibilityPoorPoorExcellentExcellentGoodExcellentExcellent

Worked example. A bank wants a new system, but users can’t clearly describe what they need and the team has never used the target technology.

  • Unclear requirements → rule out Waterfall/Parallel (Poor).
  • Unfamiliar technology → Prototyping is Poor.
  • Throwaway prototyping is Excellent on both → best fit.

⚠ Prototyping is Poor for complex or reliable systems (design gets messy) — a classic exam trap.


5. Object-Oriented Systems: Basic Characteristics

5.1 Classes & Objects

  • Class: template for objects.
  • Object (instance): an instantiation of a class.
  • Attributes: information describing the class.
  • State: the object’s attribute values and relationships at a point in time.

5.2 Methods & Messages

  • Methods: the behavior of a class.
  • Messages: information sent to an object to trigger a method (like a procedure call).

5.3 Encapsulation & Information Hiding

  • Encapsulation: combining process and data in one unit.
  • Information hiding: the internal functionality is hidden; only the interface is visible.

5.4 Inheritance

General classes (superclasses) are created; subclasses inherit their data and methods.

5.5 Polymorphism & Dynamic Binding

  • Polymorphism: the same message is interpreted differently by different classes.
  • Dynamic binding: the object’s type isn’t determined until run time (vs. static binding — decided at compile time).

Worked example.

classDiagram
    class Person {
        -name
        -address
        +insert()
    }
    class Patient {
        -insuranceCarrier
        +insert()
    }
    class Doctor {
        -specialty
        +insert()
    }
    Person <|-- Patient
    Person <|-- Doctor
  • Patient and Doctor inherit name and address from Person.
  • Sending the same message insert() to a Patient or a Doctor runs different code → polymorphism.
  • Which insert() runs is decided only when the actual object is known at run time → dynamic binding.
  • The caller only sends the message; it doesn’t know how insert() works → information hiding.

6. Object-Oriented Systems Analysis & Design (OOSAD)

OOSAD balances data and process, using UML and the Unified Process. Three characteristics:

6.1 Use-Case Driven

  • Use cases define system behavior: how users interact with the system to perform an activity (place an order, make a reservation, search…).
  • Used to identify and communicate requirements to developers.
  • Each use case focuses on one business process.

6.2 Architecture Centric

The architecture is described through three views:

ViewFocusExample
Functional (external)User’s perspective — what users can doInterface, user actions
Static (structural)Classes, attributes, methods, relationshipsCustomer with placeOrder(), updateProfile()
Dynamic (behavioral)Messages between classes and resulting behaviorSequence diagram

6.3 Iterative & Incremental

Continuous testing and refinement; the analyst understands the system better with each iteration.

6.4 Benefits

  • Break a complex system into smaller, manageable modules.
  • Work on modules individually.

7. The Unified Process (UP)

Definition

A methodology that defines when and how to use UML techniques in OOAD. Two dimensions:

  • Phases — stages of development (Inception, Elaboration, Construction, Transition).
  • Workflows — activities performed in each phase (requirements, design, coding, testing…).

Phases and workflows overlap: every workflow can be active in every phase, with varying intensity.

7.1 Phases

UP phase≈ SDLCFocus
InceptionPlanningFeasibility analyses; business modeling & requirements
ElaborationAnalysis + DesignHeavy analysis & design
ConstructionImplementationProgramming
Transition(Installation)Testing & deployment

Feasibility analyses in Inception

  • Technical — do we have the capability to build it?
  • Economic — will it provide business value?
  • Organizational — will the organization use it?

7.2 Workflows

Engineering workflowsSupporting workflows
Business modeling, Requirements, Analysis, Design, Implementation, Testing, DeploymentProject management, Configuration & change management, Environment
flowchart LR
    I[Inception] --> E[Elaboration] --> C[Construction] --> T[Transition]
    subgraph Peak workflows
      I -.- BM[Business modeling<br/>Requirements]
      E -.- AD[Analysis<br/>Design]
      C -.- IM[Implementation]
      T -.- TD[Testing<br/>Deployment]
    end

8. Unified Modeling Language (UML)

Definition

A standard language to visualize, design and document systems — the blueprint of a software system. It gives developers a common vocabulary of OO terms and diagramming techniques, and is used from analysis → design → implementation.

Why UML? Communicate clearly in teams, model complex systems visually, avoid confusion before coding, bridge business and technical understanding.

UML 2.5 has 15 diagrams in 2 groups:

8.1 Structure Diagrams

Show how the system is built — data and static relationships.

DiagramUsed to…Primary phase
ClassRelationships between classesAnalysis, Design
ObjectRelationships between actual objects (instances)Analysis, Design
PackageGroup UML elements into higher-level constructsAnalysis, Design, Implementation
DeploymentPhysical architecture; software components on hardwarePhysical Design, Implementation
ComponentPhysical relationships among software componentsPhysical Design, Implementation
Composite StructureInternal structure of a classAnalysis, Design
ProfileDevelop extensions to UML itselfNone

8.2 Behavior Diagrams

Show how the system behaves — dynamic relationships among objects.

DiagramUsed to…Primary phase
ActivityBusiness workflows, flow of a use case, or detailed method designAnalysis, Design
SequenceObject behavior in a use case — time orderingAnalysis, Design
CommunicationObject behavior in a use case — communication among objectsAnalysis, Design
Interaction OverviewOverview of flow of control of a processAnalysis, Design
TimingInteraction among objects and their state changes along a time axisAnalysis, Design
Behavioral State MachineBehavior of one classAnalysis, Design
Protocol State MachineDependencies among the interfaces of a classAnalysis, Design
Use-CaseBusiness requirements; interaction between system and environmentAnalysis

⚠ The use-case diagram is a behavior diagram, not a structure diagram. Class/object diagrams are structural.