A large Dutch campus consists of several locations with critical and non-critical functions. Amongst which a hospital with ICU units and a laboratory that performs specialist tests. Its energy system includes local renewable generation, battery storage, flexible energy demand and emergency generators with a limited fuel supply.

During a global cyber crisis an essential external energy optimisation service and remote supplier support become unavailable for an unknown period. The local installations remain operational, but due to shortage in staff only a small number of experienced operators and technicians are available to monitor, inspect and maintain them. Due to an ongoing heatwave more energy is needed to maintain the temperature in critical locations than usual.

The crisis team must continuously decide:

• which installations and locations must remain fully operational;

• which locations can be temporarily scaled down or switched off;

• where available energy, technical staff and maintenance capacity should be deployed;

• when physical inspection or intervention is required.

First design the campus’s future physical energy system. This includes deciding how energy is generated, stored, converted and distributed, how flexibility is used and which forms of redundancy are required. The design must enable further electrification while protecting critical healthcare and laboratory processes.

Design an AI-supported operational decision model that predicts the impact of scaling individual locations up or down and helps the crisis team prioritise its actions. The model should combine factors such as criticality, expected energy demand, available generation and storage, technical condition, operational risk and recovery time.

Also propose one realistic robotics application that reduces the workload or exposure of scarce technical staff, for example by remotely inspecting installations, detecting anomalies or supporting a physical intervention.

Teams must determine which decisions AI may recommend or execute, when human judgement is required and what information a human decision-maker needs to accept, adjust or overrule the recommendation.

A practical concept that can be developed within one day, consisting of:

• A design of the physical energy system covering generation, storage and distribution, enabling electrification and CO2 reduction;

• a visual decision model showing how locations and installations are prioritised;

• a simple predictive model or prototype that demonstrates the impact of scaling at least three different locations up or down;

• a proposed division of tasks between the operator, AI and robotics;

• one robotics use case that supports inspection, monitoring or intervention;

• a crisis dashboard, mock-up or decision flow showing the recommended action, expected impact, uncertainty and required human decision;

• a short demonstration using at least two contrasting crisis scenarios.

The proposal must explain the assumptions and trade-offs behind its recommendations. There is no single mathematically optimal answer: teams must use human judgement to balance safety, continuity, energy availability, technical risk and scarce personnel.