2. Requirements Definition

The speaker emphasizes that many project problems, including schedule and cost overruns, are traceable back to poorly defined, over-ambitious, or missing requirements. The case of the 1998 Mars Climate Orbiter is presented as a prime example of mission failure due to a units confusion problem, which stemmed from a failure to enforce written technical interface specifications.

In contrast, the DC-3 aircraft, designed in the 1930s, is highlighted as a success story where high-level, clear, and well-defined requirements led to a hugely successful product. The requirements were surprisingly concise, focusing on range, cruise speed, passenger capacity, and general robustness.

The video illustrates the phenomenon of 'requirements explosion,' where as systems evolve and become more complex, the number of requirements grows exponentially. From the simple 'get off the ground' for early flight to modern avionics with thousands of requirements for radar transparency, fatigue, and 'ilities' like affordability, this trend poses significant challenges for complexity management.

The importance of requirements is underscored by their inclusion in major system engineering standards like the INCOSE handbook and ISO standards. The core purpose of requirements is to set goals and constrain design within the objective space, distinguishing between 'shall' statements (hard constraints) and 'should' statements (desirable goals). An example of house design is used to clarify these concepts.

A key distinction is made between 'requirements' and 'specifications.' Requirements define what a system shall or should do, its functions, and performance (user-centric). Specifications, on the other hand, describe how a system is built, its form, materials, dimensions, and user interface details (design-centric). A microwave oven example is used to illustrate this difference.

The NASA system engineering engine emphasizes technical requirements definition as a critical step. The 'shall' statement is presented as a firm constraint, differentiating it from a 'should' (goal). Technical requirements are essential for transforming stakeholder expectations into measurable terms, providing a basis for agreement, reducing rework, estimating costs and schedules, and facilitating verification and acceptance.

The video explains that while ideally all requirements would be defined upfront, this is often impossible because lower-level requirements depend on key design decisions. The process is iterative, where high-level requirements are defined first, followed by design choices, which then enable the definition of more detailed, lower-level requirements.

Six types of requirements are outlined: functional (what the system does), performance (how well it performs), constraints (non-negotiable limits like mass or power), interface (how the system connects with others), environmental (conditions under which it operates), and 'other' (human factors, reliability, safety, often overlooked but crucial).

A good individual requirement should be clear, consistent, correct, feasible, flexible, unambiguous, and verifiable. For sets of requirements, they should exhibit absence of redundancy, consistency in terminology, completeness, and absence of direct conflict. Feasibility is particularly challenging, especially for novel projects, as it relies on the state of the art and project constraints.

Requirements decomposition involves breaking down high-level requirements into lower-level functional and performance requirements for subsystems. The terms 'allocated' and 'derived' requirements are introduced. Requirements margins management, which involves building in reserves for mass, power, etc., is crucial due to uncertainty. Historically, mass growth in aerospace projects can range from 10% to 60%. The system requirements review (SRR) acts as a critical milestone for vetting high-level requirements. The concept of ISO performance, a methodology for allocating requirements to lower-level parameters while maintaining overall system performance, is briefly introduced using a space telescope example. Finally, the video discusses the practical management of requirements, recommending document-based methods for smaller projects and specialized database tools like DOORS for larger, more complex projects with thousands of requirements.

The flow-down of requirements, ensuring that satisfying lower-level requirements guarantees satisfaction of higher-level ones, is a significant challenge. The lecture concludes by emphasizing that requirements are never 'frozen' entirely; they will be updated and refined throughout the project lifecycle. Requirements volatility, or the rate at which requirements change, is a critical factor that can significantly impact project cost and schedule, thus needing careful management.