Find what limits
the outcome.
The slowest dependency can set the pace for the entire project. Identify it before optimizing work that cannot improve overall delivery.
Systems thinking / Engineering / Delivery
The constraint sets the pace.
Complex projects are systems of dependencies. I focus on what actually limits progress - connecting engineering, people and site execution to move the whole project forward.
Complex systems · One directionConceptual industrial systems diagram
Less local optimizationBetter decisions across interfacesMore meaningful progress
01 / The thinking
See the system.
Not just the tasks.
Schedule pressure can make every activity look urgent. Systems thinking helps distinguish movement from progress - and directs attention to what affects the outcome.
Find what limits
Manage the
interfaces.
Engineering, materials, contractors, access and commissioning are connected. A gap between teams often matters more than activity within one team.
Protect progress
through the system.
A full schedule is not the same as productive flow. Priorities, sequencing and reliable handoffs should support completion - not just utilization.
Local efficiency
is not system performance.
Optimizing one discipline in isolation can create queues, rework and delays elsewhere. The objective is not to make every part look efficient; it is to help the entire project deliver reliably, safely and with fewer interruptions.
Approach inspired by Eliyahu M. Goldratt's Theory of Constraints · Not a direct quotation
02 / How I approach delivery
Focus.
Align.
Improve.
Identify the constraint
Find the decision, work front, technical issue or missing prerequisite that is holding back meaningful progress.
Make the best use of it
Remove avoidable interruptions, clarify inputs and protect scarce capacity before asking for more resources.
Align surrounding work
Sequence engineering, logistics and site activities around the real priority rather than separate local targets.
Increase capacity where needed
Escalate decisions, change the method or secure additional support when the existing system cannot deliver enough.
Reassess and repeat
Once one constraint is resolved, another may emerge. Update priorities using the actual state of the project.
03 / The delivery system
Activity is not
throughput.
Engineering, materials and site readiness can all be progressing - while delivery remains constrained. The task is to identify what governs output, then align the work around it.
Better in landscapeRotate your phone to explore the full diagram.
SYSTEM FLOW / CONCEPTUAL MODEL GOVERNING CONSTRAINT
↔ Swipe to explore the full diagram
01 / RECOGNIZE
Find the limiting factor.
Identify where delivery slows, not simply where activity is highest.
02 / ALIGN
Protect the constraint.
Coordinate design decisions, supply and site priorities around what enables progress.
03 / IMPROVE
Increase the flow.
Reduce waiting, resolve interfaces and reassess when the constraint moves.
04 / The root cause
Symptoms are
not causes.
Several problems can look unrelated yet arise from the same underlying condition. Instead of treating each symptom separately, trace the cause-and-effect relationships and test what actually drives them.
CURRENT REALITY TREE / SIMPLIFIED EXAMPLECAUSE → EFFECT
01 / POSSIBLE COMMON CAUSE
Cross-team decisions lack clear ownership
02 / SHARED MECHANISMInterfaces remain unresolved
03 / VISIBLE EFFECTSLate decisionsTeams waitingRepeated work
WHAT TO CHANGEMake decision ownership and escalation paths explicit.
01 / DIAGNOSE
Separate what is happening from what may be causing it.
02 / CHANGE
Test the causal logic, then address the shared cause rather than its visible effects.
05 / The conflict
Challenge the assumption.
Not the people.
Two teams can disagree on what to do while working toward the same goal. The conflict often survives because an untested assumption makes the choices appear mutually exclusive.
EVAPORATING CLOUD / SIMPLIFIED EXAMPLEONE GOAL · TWO NEEDS · TWO ACTIONS
A / SHARED GOALSafe, predictable delivery
B / VALID NEEDProtect work readiness
↓
D / ACTIONHold the whole workfront
C / VALID NEEDMaintain site progress
↓
D′ / ACTIONProceed despite open interfaces
COMPETING ACTIONS
ASSUMPTION TO TEST
“We must either stop everything or proceed with everything.”
BREAK THE ASSUMPTION ↗
Release verified workfronts. Hold unresolved interfaces. Both readiness and progress can be protected without treating the entire site as one decision.
The method is not about choosing a compromise between two positions. It is about understanding the needs behind both positions and finding an alternative that satisfies them.
THINKING PROCESSES / INSPIRED BY E. M. GOLDRATT06 / Applied in practice
Engineering meets
execution.
A systems view is most useful when combined with hands-on technical understanding and direct experience of work on site.
Interfaces &
dependencies
Aligning disciplines, design information, technical decisions and work prerequisites across project teams.
Site execution
& control
Supervision, contractor coordination, work sequencing, progress tracking, safety, quality and handover readiness.
Technical systems
& integration
Electrical installations, instrumentation, automation, BMS interfaces and commissioning documentation.
Project environments
Industrial facilities & pharmaceuticals / Shipbuilding & marine / International construction