Research pick: Emergency AI hops to it - "Intelligent decision system for urban emergency management based on combined deep learning and optimisation algorithm"

A hybrid AI, artificial intelligence, system that models itself on grasshopper behaviour could be used to help with the allocation of medical resources, transport, and power during an urban emergency, according to research in the International Journal of Environmental Technology and Management. Tests with the new hybrid model on historical emergency datasets show it to be highly effective

The team's LSTM-GOA system combines a long short-term memory neural network and the so-called grasshopper optimisation algorithm (GOA). LSTM is a machine-learning tool that can identify patterns in data as they change over time. GOA is an optimisation technique that searches for better solutions to complex problems based on how a swarm of grasshoppers forage and feed. In this hybrid approach, the GOA is used to tune the behaviour of the LSTM so that it makes better scheduling decisions in order to find the most appropriate solution.

The researchers compared the model with conventional rule-based approaches and other optimisation methods. The team was able to improve prediction accuracy for medical resource demand by almost 70 per cent. The system can respond to otherwise unpredictable spikes in demand following natural disasters, public health incidents, and major traffic accidents. Moreover, it can work with noisy or incomplete data sets.

The work points the way to the broad use of predictive AI in managing resources in an emergency. The team explains that city authorities could use historical data to anticipate pressures across several public services and adjust allocations accordingly. This would be more effective than relying on fixed rules or responding to changing demands after the fact.

Cai, X., Qiu, J., Cao, H., Wu, F. and Mei, X. (2026) ‘Intelligent decision system for urban emergency management based on combined deep learning and optimisation algorithm’, Int. J. Environmental Technology and Management, Vol. 29, No. 7, pp.59–85.

News media may use this press release as source material, in whole or in part, provided the content is not materially misrepresented. A link back to the original article is appreciated.