Question:

The topic for group discussion for the week was enquiry into large systems and to explore why some systems are large? To analyse the question posed by Schwartz rather incisively

“I must admit that I have frequently asked myself whether these systems require many people because they are large, or whether they are large because they have many people”.

Case Study:

Large Systems P41-43 of NATO Report on Software Engineering Techniques

Dissection Summary:

This NATO case study deals with the analysis of a project tasked to build a large computerized aircraft surveillance system. The enquiry highlighted the complexities faced in managing such projects which involves large number of people working together to build a complex system. The following points can be deduced from the case study

1. A successful prototype does not mean that a scaled up, fully functional system will also be successful.  In this case a prototype included only a limited number of radar inputs & tracking only a few aircrafts, but scaling it up and building a massive surveillance system covering whole country of the size of US brings to fore the its own set of complexities.  The points below are relevant for any large system project
   1. Managing Scope: The difficulties in defining scope of the system that meets at least basic user requirements. During prototyping stage the scope is generally quite small and can be easily defined.
   2. Project Management: Meeting deadlines and avoiding project slippages. This also requires managing personnel & team equations with different project teams. During prototyping stage the project management can be very simplistic as the team size and scope is small.
   3. Supervising infrastructure related complexities: Bigger & complex infrastructure requires specific skills to manage it.
   4. Resource Management:  Difficulties in hiring and managing resources for project implementation. How to recruit people with right skill-sets and required training to support the new system?
   5. Building Support Systems: Generally big systems have numerous touch points with other partner systems; planning & building these support systems also bring added scoping & planning complexities. These support systems are non-existent during prototyping stage.
2. Managing large system projects require a number of attributes to be successful. Few of them are as following
   1. Control: A large system project need strict control on different project management cycles including
      1. Scoping
      2. Analysis & Design
      3. Build
      4. Testing
      5. Implementation
      6. Also control is imperative in Configuration Management & Documentation for a successful system build.
   2. Flexibility: A large system project need flexibility to deal with
      1. Scope changes
      2. Technology changes
      3. Procedural changes
      4. Direction changes
      5. Timeline changes
   3. People: People management need to be managed properly
      1. Training to bring people up to the skill level required
      2. Measurement techniques to asses capability levels
      3. Roles & responsibilities of the key people who need to work across teams
   4. Management: Project management should take care of
      1. Planning
      2. Decision making
      3. Team structure
      4. Negotiation
      5. Motivation levels

Critique:

Though the case study raises an important question and is very well written, the underlying message is now a bit out-dated. This case study would have been very pertinent at the time it was written when such large systems were evolving and there was no precedence to build such large systems. Since then the learning from big projects involving large systems have made big organisations aware of the various pitfalls in managing such projects. The new project methodologies like RAD, XP and UP etc. successfully leverage latest tools to build large systems. These latest tools also help put enough controls in Software configuration management & documentation aspect of project management.

Another big difference is that now most of the project tasks can be outsourced to offshore companies easing the resource management aspect of such big projects. The offshore model of software development was non-existent at the time this case study was written. Also now the offshore companies have become quite matured and are specialized in particular domain & technologies and can be relied upon for the successfully delivery of the various sub components of large systems.

The article does not put enough emphasis on the design phase where most of the projects go wrong.  For large systems the importance of design phase increases manifold due to the complex nature of such systems. The importance of dissecting the proposed system functionality into well defined requirements cannot be understated.

The case study rightly emphasises the importance of control, flexibility, people & management aspects of managing large system projects and these are relevant in today’s context also.