Photo: Torbjørn Kjosvold / Forsvaret
Training Transformation
Cold Response 2026: Training a Division with Live-Constructive Integration
How Norway designed and executed the first deliberate live-constructive architecture to train its full divisional structure — and what it means for the future of collective training at scale.
This paper was submitted as a technical paper to ITEC 2026 and served as the basis for the author’s contribution to the Training Transformation Panel on 16 April 2026. It is published here in full as the author’s own work. The accompanying presentation is available under Conference Papers on the Insights page.
The Strategic Requirement: Why Division-Scale Training Has Been So Hard
The Norwegian Army operates within a strategic environment that demands readiness at the divisional scale while imposing constraints that make traditional full-force collective training increasingly impractical. Two brigades form the operational core of the 6th Division. Both cannot always train simultaneously. Operational commitments pull units. Budget limits the duration and scale of live deployments. Arctic terrain concentrates training areas in a small number of locations.
The practical result is that collective training at the division level — the echelon where multi-domain integration, logistics, fires synchronisation, and operational planning converge — has been the most difficult level to train adequately.
Cold Response 2026 posed a specific problem: the Chief of the Army required training for the entire divisional structure — both brigades, the division headquarters, and supporting formations — in a realistic and complex operational scenario. One brigade was deployable. The other was operationally committed. The decision to solve this through live-constructive integration was not primarily a technological one. It was a training design decision.
The Cold Response 2026 Architecture
Distributed Design: Three Nodes, One Operational Picture
The exercise operated across three geographically separated nodes, each performing a distinct function in the integrated training architecture.
Brigade North and allied forces conducted live instrumented manoeuvre. Home Guard units and OPFOR operated under full instrumentation. The kinetic core of the exercise.
Brigade Finnmark’s HQ operated from its actual command post, commanding constructive forces in JCATS — geographically separated by hundreds of kilometres but fully integrated in the operational picture.
JCATS support staff modelled all constructive forces. The SIMSC2 Gateway translated simulation outputs into C4IS messages. Constructive forces appeared in the divisional COP on equal terms with live forces.
Technical Components
The core technical stack comprised four integrated components: JCATS at Camp Rena modelling all constructive forces; the Norwegian Army C4IS architecture receiving and displaying all force positions and tactical data; the SIMSC2 Gateway translating JCATS outputs into C4IS-compatible messages in real time; and a secure communications backbone connecting all three nodes.
Live instrumentation added further complexity. Available instrumentation stock was prioritised for manoeuvre units — a recurring Norwegian constraint the ongoing P5066 FACTS procurement is intended to resolve. The German Mountain Ranger battalion operated the AGDUS instrumentation system, with interoperability maintained through an HLA interface, resulting in two separate data-acquisition networks across the live exercise area.
The integration principle was direct and consequential: simulation did not run in a separate training environment. It fed the Army’s operational C4IS systems. Constructive units appeared in the same operational interfaces as live instrumented forces. For the division staff and Brigade Finnmark headquarters, the picture was operationally complete.
What the Exercise Achieved
Divisional Staff Training at Full Structural Scale
The primary training objective was achieved: the 6th Division headquarters trained against a complex multi-domain scenario with its full divisional structure engaged. Fires, logistics, intelligence, and manoeuvre were coordinated across the divisional battlespace. The command staff encountered the full cognitive load of divisional command — not a reduced-scale approximation. Divisional command is qualitatively different from brigade command, and training on a scenario where one brigade is absent does not adequately prepare that staff for the complexity of actual divisional operations.
Brigade Finnmark: Exercising a Future Structure
Brigade Finnmark’s participation produced a training outcome that would not have been achievable through any live exercise arrangement. The brigade staff commanded from its actual operational command post, using its actual equipment and command systems, against a constructive force that responded to its orders in real time. The exercise environment was operationally realistic at the command level — which is where the training value for a brigade headquarters is concentrated. The brigade was also exercising a future structure: formations and capabilities that exist in planning and doctrine but are not yet fully fielded. Constructive simulation accommodates that; the order of battle in JCATS can represent the planned force structure, not just the current one.
Operational Continuity Maintained
Brigade Finnmark maintained its operational commitments throughout the exercise period. Its units did not redeploy to a training area. Its staff did not leave its area of responsibility. The Army was able to train its entire structure while sustaining its operational posture — a combination that a purely live-exercise design cannot deliver.
C4IS Integration: The Operational Distinction
The integration of constructive simulation into the Army’s operational C4IS architecture was the technical feature that made the exercise operationally meaningful rather than merely technically interesting. Simulated forces did not appear in a separate simulation display. They appeared in the same operational systems that commanders use in actual operations — using the same interfaces, reporting formats, and situational awareness tools. This is the transition that separates simulation-supported training from integrated live-constructive training. The cognitive transfer from training to operations is higher when the training environment uses operational tools rather than simulation-specific displays.
Structured Observation and Evaluation
Cold Response 2026 fielded a structured observation and evaluation capability through the Exonaut exercise management system, integrated with the Norwegian Army’s Fis-B command information system. Observer teams registered observations via mobile app and directly within Fis-B throughout the exercise. Two structural constraints limited its full potential: exercise objectives were set predominantly during or after the start of exercise rather than in the planning phase, and observer-to-unit ratios at the divisional and brigade staff levels were thin. These are planning and resourcing decisions — not Exonaut limitations.
Friction Points Exposed
Cold Response 2026 succeeded. It also exposed structural friction that must be understood if the distributed live-constructive model is to become a repeatable institutional capability. These are not Norwegian peculiarities — they reflect alliance-wide structural challenges.
Simulation was integrated into exercise planning later than optimal. The consequence was compressed preparation time for the simulation team, limited opportunity to align constructive force behaviour with exercise objectives, and reduced familiarisation time for Brigade Finnmark. The fix is straightforward in principle: simulation must be embedded in exercise planning from the initial concept stage, not introduced at the detailed planning phase.
JTLS-GO at the Joint Warfare Centre runs on a national/NATO Secret network. The Norwegian JCATS system is accredited only for national Restricted/NATO Restricted. This classification gap is one of the principal reasons the two constructive environments could not be technically federated. Synchronisation between JTLS-GO and JCATS depended entirely on procedural coordination — a dedicated liaison officer positioned at the Joint Warfare Centre. The solution worked, but it is not scalable. The defence training industry needs to develop working integration solutions for JTLS-GO and tactical constructive simulations at the classification levels where operational planning actually occurs.
Sustaining realistic constructive simulation of two-brigade operations over an extended exercise period requires substantial JCATS operator capacity. At the tactical CAX layer, each battalion in Brigade Finnmark was supported by a single NCO/Officer and two conscripts as JCATS operators. This is simultaneously a technology problem and a planning problem. Modern constructive simulation platforms with genuinely AI-assisted tactical behaviour could significantly reduce the operator-to-entity ratio — enabling the same training output with smaller support teams. On the planning side, systematic investment in operator training and a standing trained cadre would reduce the cost of each exercise.
The integration architecture assembled for Cold Response 2026 was built for the exercise. The SIMSC2 Gateway configuration, C4IS integration testing, and simulation-to-C2 data mapping required dedicated work, drawing on experience from previous exercises but not from a standing institutional baseline. The integration work done once should not need to be redone for the next exercise. A standing federation protocol — maintaining interface configurations, validating data mappings, and testing C4IS integration — would dramatically reduce the preparation cost of each subsequent exercise.
Implications for Army Training Transformation
From Proof of Concept to Standing Capability
Cold Response 2026 proved the concept. The question that follows is institutional: what would it take to make this the standard model rather than the exceptional one?
Three changes are required, and none of them is primarily technical. First, simulation must be treated as a first-tier element of exercise design — integrated from the initial concept stage, with the simulation design brief on par in priority with the live manoeuvre design brief. Second, the Army needs a modernised constructive simulation capability: JCATS has served well, but its operator-intensity, the lack of an AI model, and incomplete modern multi-domain representation are constraints that become more limiting as the training ambition expands. Third, the integration architecture needs to become a standing organisational asset — the interface configurations, data mappings, and tested C4IS integration developed for Cold Response 2026 represent institutional knowledge that should be maintained as a persistent, documented, and regularly exercised architecture.
The Deeper Transformation
Cold Response 2026 represents more than an innovative exercise architecture. It represents a conceptual shift in how collective training at the divisional scale is understood. The traditional model positions live manoeuvre as the primary training medium, with simulation as a support function. The distributed live-constructive model repositions simulation as a co-equal training layer — one that extends the operational picture beyond what live forces can occupy, enables training of operationally committed units, and provides the constructive environment in which divisional command complexity is generated.
This is the shift that matters for the defence training industry. The demand signal is clear: armies need to train at a divisional scale, under realistic constraints, with increasing frequency. The required supply response is not more elaborate live-exercise infrastructure. It is better integrated, more capable, and more accessible constructive simulation — integrated into operational C2 systems, supported by AI that reduces operator load, and maintained as a standing component of the Army training architecture.
Alliance Implications
The friction points exposed at Cold Response 2026 reflect structural challenges that every NATO army faces when it attempts to integrate live and constructive training at scale. NATO’s Distributed Synthetic Training (DST) initiative — which reached a binding Memorandum of Understanding among 13 Allies at NATO Headquarters in Brussels on 15 October 2025 — is building the alliance-level framework for a persistent, multinational, constructive training architecture. Cold Response 2026 is an early operational data point for that framework: it shows what distributed training at a divisional scale looks like in practice, and it provides the operational evidence base for the institutional decisions the DST initiative requires.
Acknowledgements
The author wishes to express sincere gratitude to the dedicated people working at the Norwegian Army Training and Simulation Centre, who made the live-constructive integration during exercise Cold Response 2026 possible, and without whom this paper would never have been written — and whom it has been an honour to work alongside and lead for the last eight years.
Specific thanks are due to the following for their contribution during and after the exercise in providing information on the conduct and outcomes:
- Lt Col Hans Petter Gretland — successor as Head of the Norwegian Army Training and Simulation Centre (NATSC)
- Lt Col Nicholas Gran — Head of Instrumentation during the exercise
- Lt Col (Ret.) Trond Martin Flatemo — Senior Staff Officer at NATSC and responsible for planning and implementing Exonaut for use during the exercise
The author used Claude (Anthropic) as a research and editorial tool in preparing this paper. Claude assisted with structuring the paper’s argument, synthesising case study material, and iterative drafting. All operational case study material, factual claims, professional judgements, and conclusions are the author’s own, drawn from direct experience. The author reviewed, corrected, and revised all AI-assisted content throughout the drafting process. Any errors are the responsibility of the author.
Glossary of Key Terms
| AGDUS | Ausbildungs-Gefechts-Diagnosesystem — the German Army’s instrumented training system, used at the national Combat Training Centre in Güz, Germany. |
| C4IS | Command, Control, Communications, Computers, Intelligence, and Surveillance — the integrated information systems through which military headquarters exercise command and maintain situational awareness. |
| CAX | Computer-Assisted Exercise — a training event conducted primarily or entirely through constructive simulation, without live forces in the field. |
| DST | Distributed Synthetic Training — NATO initiative for a persistent, multinational constructive training architecture. MoU signed by 13 Allies in Brussels, 15 October 2025. |
| HLA | High Level Architecture — IEEE standard (IEEE 1516) for distributed simulation interoperability, enabling data exchange between incompatible simulation systems. |
| JCATS | Joint Conflict and Tactical Simulation — a constructive simulation system developed by Lawrence Livermore National Laboratory, widely used by NATO armies for battalion- to corps-level training. |
| JTLS-GO | Joint Theater Level Simulation — Global Operations. Operational- and strategic-level constructive simulation used at the Joint Warfare Centre. |
| LVC | Live, Virtual, Constructive — the three domains of military simulation: live forces with instrumentation; virtual forces in simulators; constructive forces in computer-generated environments. |
| NATSC | Norwegian Army Training and Simulation Centre — the national centre responsible for simulation-based training support to the Norwegian Army. |
| OPFOR | Opposing Force — the simulated or designated adversary force in a training exercise, tasked with providing a realistic and challenging threat to the training audience. |
| P5066 FACTS | Future Army Combat Training System — Norwegian Ministry of Defence procurement project to reacquire and expand the Norwegian Army’s Combat Training Centre instrumentation capability. |
| SIMSC2 | Simulation-to-Command-and-Control Gateway — middleware that translates simulation outputs into operational C2 data formats, enabling constructive forces to appear in live command systems. |
Lieutenant Colonel (Ret.) Ken-Tore Saade Eriksen retired from the Norwegian Army after 38 years of service. He served as Head of the Norwegian Army Training and Simulation Centre from 2018 to 2026, with national-level responsibility for integrating Army-wide simulation capabilities.
He has been a member of two NATO Modelling and Simulation Group (NMSG) sub-working groups: Simulation for Training and Operations – Land (STOG-L) and Urban Combat Advanced Training Technologies – Live Simulation Standards (UCATT-LSS). He was the author of the keynote address by the Chief of the Norwegian Army at ITEC 2025, and a panel speaker at ITEC 2025 and ITEC 2026.
He is the founder of Eriksen LiveSim Consult, an advisory firm specialising in the integration of synthetic training and defence training modernisation, and currently serves as Business Development Director for Nordics & Baltics at BAE Systems OneArc.
