A Modular Compositional Approach to Engineering Complex Systems
summary
Argues for a practical modular compositional modeling method that makes system models self-contained and recursively composable, tying roles to concrete assumptions and bindings so properties can be checked across levels. Cites a radio‑telescope program with 4 years, 500 people, 10 subsystems where dozens of toolsets (SysML, MATLAB, CAD, network models) never yielded a unified machine‑readable model. Shows a TCS AI pilot that delivered 3% production gain refused deployment and outlines runtime monitoring, commitment envelopes and closed‑loop model updates.
AI-written summary · about
chapters (5)
- 0:00Start
- 4:04What is the Nature of Digital?
- 11:11Cynefin Perspective
- 14:13Is Our Engineering Digital?
- 22:16Reframing Compositionality
related talks
52mGood Fences Make Good Neighbours
51mModularization: Foundation of Microservices and Monoliths
01mMark Richards at DDD Europe
48mGood Fences Make Good Neighbours
46mFrom Chaos to Cohesion: Building Modular Monoliths in the Real World
43mGood Fences Make Good Neighbours
31mModularity in open systems
01h 24mA new modularity for software