Doubts about the reliability of autonomous killing machines were flagged in the 1984 movie Terminator
Experimental feature

Listen to this article

00:00
00:00
Experimental feature
or

The warning is stark. Humans could be left “utterly defenceless” against robotic weapons systems, according to Professor Stuart Russell, specialist in artificial intelligence at the University of California, Berkeley.

Writing in Nature magazine, Prof Russell describes a bleak future where swarms of miniature drones could independently identify and attack targets without the need of human intervention. Using technology developed for self-driving cars, these flying nanobots could be deployed not just on the battlefield but on search-and-destroy missions in heavily built-up urban areas.

There is no inevitability about this scenario. But as governments prepare to equip their military forces for the next generation of warfare — in particular with unmanned aerial systems — the debate over how far to push technology is beginning to test industry, governments and specialists such as Prof Russell.

According to him, the components exist to build a combat drone smart enough to make lethal decisions on its own. “They just need to be combined,” he says.

However, putting these capabilities together into one vehicle remains a challenging task. Other industry experts estimate that the technology to build a fully autonomous weaponised drone is still at least 20 years away.

Moreover, development of drones with sufficiently advanced artificial intelligence to make lethal decisions without human intervention remains highly contentious — in some countries, at least.

Douglas Barrie, analyst with the defence think-tank The International Institute for Strategic Studies (IISS) in London, believes that public opposition to “killerbots”, as the media labels them, means there will always be a human making attack decisions for drones operated by western militaries, no matter how smart they might become.

But that does not mean the technology to develop such artificial intelligence can be ignored. “The west has to be cognisant that it might not see the same levels of restraint in other countries,” he says. Even the UK, which has far stricter rules of engagement than the US, recently opposed an outright ban on the development of lethal autonomous weapons systems (Laws).

“Of course we’re thinking about this technology,” says one European defence industry executive. “It is not one we are particularly working on it, but if others go autonomous we’ll need to do it too.”

As the debate rages, military forces around the world are increasing their reliance on unmanned aerial vehicles (UAVs). The aim is to reduce cost and casualties, while increasing effectiveness — although some forces are discovering that drones can be more costly to operate than expected.

From just 162 in 2004, the US Department of Defense now operates more than 8,000 drones — roughly 40 per cent of the air fleet — including aircraft for intelligence, surveillance and reconnaissance (ISR), as well as attack missions. A decade ago, just three countries — the US, Russia, and Israel — deployed UAVs. Today, the militaries of more than 70 countries use remotely piloted aircraft, according to the IISS.

Europe has mainly used technology developed in the US and Israel for its drones. But recent initiatives suggest that the pressure for European capability in the domain of armed manned aircraft is intensifying as global tensions mount and forces face tighter budget constraints. A new generation of combat aircraft is also necessary if Europe is to retain vital skills in its aerospace industry.

Last year’s agreement between France and the UK to work on a joint feasibility study for a future air combat system was a big step forward. The UK’s BAE Systems and Dassault Aviation of France were awarded a £120m contract in November by the UK and French governments, to identify the key technologies required for a new-generation combat drone.

A decision will be taken next year on whether the companies will go on to build a demonstrator, at which point support could be invited from other European nations to help reduce the development costs.

At that stage, BAE and Dassault will also have a clear idea of what the technology should be able to deliver by the time a new UAV comes into service — which is likely to be about 2035, when the current generation of Typhoon and Tornado manned fighter jets needs to be replaced. “We are not just working on the airframe,” the BAE executive says. “This is also about the system and the engine.”

Among the biggest challenges now, according to Pierrick Lerey, director of strategy for ISR at Thales Airborne Systems, is developing multifunction sensors. Rotating antennas, each dedicated to one technology, will be replaced by a single set of flat panels providing multifunctional capabilities such as radar, electronic warfare and communications. In the future, those panels will be shaped to the drone’s curves in a ‘smart skin’ that gives 360-degree coverage, and also contributes to the aircraft’s stealth.

The industry is also working on vital ‘sense and avoid’ systems, which will allow unmanned aircraft to fly alongside other drones or manned jets without colliding. This month, Alenia Aermacchi of Italy claimed a breakthrough in collision avoidance when it successfully flew its Sky-Y drone alongside one of its C-27J transport aircraft.

But there is still a long way to go before Europe’s new-generation combat aircraft take to the skies. Debates are raging over rival technologies, for example in sense and avoid.

Both France and the UK continue to pursue their own demonstrator projects at the same time as working together on their joint feasibility study. The UK has its Taranis stealth fighter, hidden away in a secret Lancashire location, while France is leading the seven-nation Neuron project.

“Experimental programmes such as Taranis and Neuron are important to help decide the best choices for our needs in 2030,” says Mr Lerey. “To have the technology ready, we need Taranis and Neuron to mature.”

More than that, the companies need political conviction to support what will be a long development process costing billions of euros. The evidence so far, says one defence industry executive, is that France and the UK are determined to press ahead with a joint programme, given the prohibitive costs of going it alone.

But Britain’s forthcoming referendum on EU membership could weaken what conviction does exist, says Mr Barrie. “France would like to make this a truly European project,” he adds. “That would be harder to do if the UK has pulled stumps on membership.”

Copyright The Financial Times Limited 2017. All rights reserved.
myFT

Follow the topics mentioned in this article

Follow the authors of this article