Aid agencies test AI tools to reach people in dangerous disaster zones

GENEVA: Humanitarian organisations are exploring how artificial intelligence can help them deliver assistance more quickly and safely, with new tools on display this week at the AI for Good summit in Geneva.

According to AFP, the technology being examined ranges from remotely operated aid vehicles for hazardous terrain to systems that analyse mobile phone and satellite data to improve disaster response. Aid agencies have also been warned about the risks linked to AI, including the protection of highly sensitive information and the spread of misinformation affecting both their operations and the people they assist.

WFP prepares remote-controlled vehicle trial

One of the most prominent exhibits at the summit was a large white Sherp vehicle built in Ukraine and fitted with cameras, sensors and a drone landing pad. The amphibious vehicle can float, move through swamps and flooded rivers on oversized wheels, and cross obstacles up to one metre high.

The UN World Food Programme is preparing to field-test an AI-assisted version of the vehicle that can be guided remotely across especially risky and hard-to-reach areas. Bernhard Kowatsch, head of WFP’s global accelerator and ventures innovation division, said the technology could significantly expand access to people in need.

“I think this could be a game-changer,”

Kowatsch told AFP. He added:

“should allow us essentially to reach people that otherwise never would have been reachable”

WFP already uses Sherp vehicles with drivers in Sudan, South Sudan and Uganda to transport aid. After suffering repeated driver fatalities, the agency asked the German Aerospace Center, or DLR, to help adapt the vehicles with AI and other systems so they can be controlled remotely in the most dangerous stretches.

The plan is to place a control room inside a shipping container in a secure area, from where a human operator would steer the vehicle through the final and most hazardous part of the route. Armin Wedler, who coordinates DLR’s Autonomous Humanitarian Emergency Aid Devices project, told AFP that testing has already taken place in Germany and that field trials are due to begin in Uganda in 2028.

Standing beside the 2.8-metre-high vehicle, Wedler said the team had relied on

“remote-control technologies which are based on mathematics and old-school … research”

but added:

“We would not be able to process everything without using also AI.”

He said the vehicle could technically be made fully autonomous, but stressed that humanitarian operations require human oversight.

“We have to have a human in the loop.”

He added that such environments are far removed from ordinary road conditions:

“We’re not talking about driving on clear streets with clear lanes. There are no streets,”

and said there are also situations in which hungry crowds suddenly surround aid trucks.

“There’s no AI autonomous algorithms ever capable to handle that safely.”

Legal and disaster-response applications

Other humanitarian applications at the summit were less visible but focused on improving the speed and efficiency of aid work. The UN refugee agency presented a Legal Virtual AI Assistant designed for lawyers and legal officers working with refugees. The system is intended to help them quickly identify rights available under different national legal frameworks.

Rebeca Moreno Jimenez, lead data scientist at UNHCR’s Innovation Service, told AFP that being able to prepare cases faster and more effectively can be

“life-saving for many refugees”

.

Another UN-backed initiative, Data Insights for Social and Humanitarian Action, or DISHA, works with private-sector partners including Google and McKinsey to supply humanitarian groups with data and AI models aimed at improving disaster response.

One DISHA project uses AI to examine anonymised mobile phone data in order to identify large-scale population movement during emergencies and show where people are fleeing. Another applies AI to quickly compare satellite images taken before and after disasters, including last month’s earthquakes in Venezuela, to assess damage to buildings.

DISHA product lead Andreas Kortis told AFP that the purpose is to provide humanitarians with

“accurate information early enough to make better decisions (and) avoid going to the wrong place when there are people who need you somewhere else”

.