The Secret of Apollo: Systems Management in American and European Space Programs
By Stephen B Johnson (2002)
Both Servan-Schreiber and McNamara believed that their societies needed more and better management. According to McNamara, “Some critics today worry that our democratic, free societies are becoming overmanaged. I would argue that the opposite is true. As paradoxical as it may sound, the real threat to democracy comes not from overmanagement, but from undermanagement. To undermanage reality is not to keep it free. It is simply to let some force other than reason shape reality.”
My Notes
Management History
Scientific Management
Taylor, early twentieth century.
Knowledge Workers
Drucker, 1940s.
Management by objectives:
Managers and professionals jointly negotiated the objectives for the agency or firm on the one hand and for the individuals on the other, each worker agreeing to the terms.
Concurrency
Research and development in parallel to manufacturing, testing and production. Cold War, 1940s and 1950s.
- Unpredictable costs
- Inconsistent results
- Rework
Systems Management
Manhattan Project, NASA, etc. 1950s.
- Speed (government/defence)
- Novelty (science)
- Dependability (engineering)
- Predictable costs (management)
Very large and very complex projects, e.g., Apollo:
- $19,000,000,000 budget
- 300,000 people
- 20,000 companies
- 200 universities
- 80 countries
💡 Know the context in which you’re operating.
For most technologies, building a few prototypes and performing detailed tests with them before manufacturing is feasible and sensible. Lack of coordination and planning (each of which costs a great deal) can be overcome through prototype testing and redesign of the prototype. This option is not available for most space systems, because they never return.
Features of Systems Management
Origins at Caltech, MIT, Bell Labs.
Slower development but increased reliability and cost predictability.
- Strong focus on compatibility/integration and quality/testing
- Configuration/change management
- Phased planning
- Formal design reviews
- Centralised planning and control
- Master schedules, multiple levels, continually updated
- Regular reports
- Identify, escalate and address problems
- Matrix organisation
- Design freezes
- Strong organisation skills
- Strong leadership
1x component = 90% reliability 2x components = 90% x 90% = 81%
In this way, engineers estimated the decrease in reliability as they added electrical components. With calculations such as these, engineers determined that adding a second parallel “string” of components significantly improved missile reliability.
💡 Can this kind of simple maths be extrapolated to e.g., number of features?
Finding the Right Balance
Continual back and forth:
- Processes changed as projects proceeded
- Teams often started with less process than they end up using
- E.g., when reliability is an issue, slow down, do more testing
- Inevitable push back, but teams would learn the “hard way”
As stated by one of the engineers on Gemini and Apollo, “You would see people who would try to build empires, who would try to be obstructionist, and they would be just absolutely steamrolled by this team. I saw it time and time again where there was this intense feeling of teamwork. It wasn’t always smooth, but it was like, ‘We’ve got a common goal.’”
ELDO
A “fatally flawed management structure that was almost the exact antithesis of systems management in the United States.”
- Project manager’s authority was virtually nil
- Interfaces not given enough attention
- QA randomly applied
The Germans formed four committees: a government committee to investigate the failure, a committee to investigate the rest of the design, an internal committee of the contractor ASAT, and a committee to oversee and coordinate the other three committees. 😂
Control Systems and Project Management
Hoernke’s analogy of an engineering control system and project management:
- Process (the actual project development) feeds into the sensor
- Sensor (data collection) feeds actual performance into the collator
- Collator (controller) compares actual and expected
- Memory (files) records actual and expected
- Effector (manager) modifies the process
Lean?
Lean serves a different purpose:
Whereas Japanese managers and engineers focused on the reliability of mass production processes, their American counterparts concentrated on the reliability of R&D.