Integrated Air Defence: Where to Put the Command Line
The three services complete a large annual exercise in a desert sector. Post-exercise analysis finds four separate air-defence coordination cells, three incompatible data links, and a ninety-second delay between an unmanned aircraft being detected by an Army radar and a fighter being scrambled. A proposal goes to the Chiefs of Staff Committee to create an integrated air-defence command owning all surface-to-air assets, radars and the airspace control apparatus, with the Air Force as lead service. The Air Force supports integrating airspace management but argues that absorbing Army air-defence regiments would strip the corps of organic protection. The Army argues that its guns and radars are under-used because they sit outside a single air picture. The Navy objects that its shipborne systems are inseparable from the ship. Separately, the Ministry wants any new command created without fresh expenditure in the current budget cycle.
1. Using explicit criteria, decide whether a fully integrated air-defence command should be created now.
Read the worked answer
Criteria to apply
- Is the domain single and indivisible? 2. Are the assets scarce and shared? 3. Does fragmentation cost decision time? 4. Does integration improve accountability? 5. Does it destroy service-specific expertise? 6. What does it cost?
Assessment
Air defence scores high on criteria 1 to 4: there is one airspace, the sensors and shooters are scarce, the ninety-second delay is a direct cost of fragmentation, and four coordination cells mean nobody is accountable for the engagement decision.
It scores poorly on criterion 5 if integration is implemented as transfer of ownership, because air-defence regiments train with the formations they protect and know their terrain and movement cycles.
Conclusion - integrate in layers, not in one step
- Layer 1, now, near zero cost: one air picture, one engagement-authority matrix, common identification procedures, and a single joint cell during operations. This alone removes most of the delay.
- Layer 2, medium term: operational control of designated Army air-defence units by the joint commander inside a defined battlespace, with the Army retaining administrative control, training and posting.
- Layer 3, only if Layer 2 proves insufficient: full transfer of ownership.
Judgement: the reported symptom - delay - is a command-and-control problem, not a property-rights problem. Creating a headquarters before the data architecture exists produces a command on paper with no faster kill chain, which is the classic failure mode of integration driven by org-chart logic.
2. Is the Ministry's 'no fresh expenditure' constraint a valid basis for design, or should it be resisted?
Read the worked answer
What integration actually costs
Most of the early gains are procedural: a common engagement-authority matrix, standard identification friend-or-foe procedures, joint manning of a cell, and shared rules for who may fire at what. These cost training time and doctrine-writing effort, not capital.
The expensive element is the network layer - data links, message formats, and sensor fusion that lets an Army radar hand a track to an Air Force fighter without a voice call. A zero-cost mandate can therefore deliver Layer 1 and nothing more.
Why the constraint must be challenged
A headquarters created without a supporting architecture is worse than no headquarters: it consumes senior officers, creates a new coordination layer, and gives the appearance of integration while the ninety-second delay persists. The constraint is also a false economy, because incompatible links are paid for repeatedly in every exercise and in every procurement of a new sensor.
Recommended handling
- Accept the constraint for the current cycle: implement Layer 1 fully, measure the delay again in the next exercise.
- Present the network layer as a funded line item in the following cycle with a measured baseline - 'ninety seconds to twelve seconds' is a budget argument, not a sentiment.
- Sequence procurement so that every new radar and launcher bought from now on must meet a common data standard. This is cheap now and prevents the problem recurring.
Judgement: the correct response to a resource constraint is to sequence, not to pretend. Accepting the constraint while promising integration would produce the appearance of reform and none of its effects.
3. How would you meet the Army's concern about losing organic protection and the Navy's objection about shipborne systems?
Read the worked answer
The governing principle
Command authority should sit at the level where the decision must be made, and no higher. Air defence of a moving armoured formation against a fast, low threat is a formation-level decision measured in seconds. Air defence of the wider airspace is a theatre-level decision. The design must therefore split the problem by battlespace, not by service.
Meeting the Army's concern
- Retain regiments under the Army for administration, training, posting and discipline.
- Place them under operational control of the joint commander only within a defined battlespace and only for the duration of operations.
- Publish an allocation matrix in advance that guarantees the corps a specified level of protective fire, so 'joint' does not mean 'diverted'.
- Give the corps commander a formal right of appeal to the joint commander when allocation falls below the guaranteed level.
Meeting the Navy's objection
The Navy is right that a shipborne system is part of the weapon system and the ship's survival; a ship cannot hand its close-in defence to a shore-based commander. So the ship keeps self-defence authority, while contributing tracks to the common picture and accepting the joint commander's identification and weapons-tightening instructions in a designated area.
Reasoning: the risk in all joint arrangements is a command with responsibility but no authority - coordination fora that cannot order anything. The charter must specify who may delegate engagement authority, in what area, and who is accountable for a blue-on-blue. Without that, integration adds a layer and solves nothing.