print preview

Significance of increasingly autonomous systems in a military context

More autonomous robots and weapon systems provide numerous military advantages and harbour the potential to revolutionise warfare. However, the associated challenges and risks should not be underestimated. They are still not sufficient today to bring about a clear international regulation on the use of autonomous weapon systems. armasuisse S+T supports Switzerland in the discussion surrounding the use of autonomous systems.

Pascal Vörös, Swiss Drone and Robotics Centre SDRC, armasuisse Science and Technology in cooperation with the area Armed Forces Development Programme/Armed Forces Planning Division

A man and a girl are standing in front of the futuristic looking main building of armasuisse S+T in Thun. Standing in front of them is the walking robot ANYmal while drones hover above them in the sky.

In 2018, Elon Musk and 160 organisations from research and development, as well as 2400 other experts in the campaign «Stop Killer Robots» appealed against the construction of autonomous weapon systems. The subject of such a threat thus, for the first time, raised the awareness of many in the population at large. 

As a part of the technology centre of the DDPS, armasuisse Science and Technology, the Swiss Drone and Robotics Centre (SDRC) deals intensively with the current and future applications of unmanned, autonomous systems. In this article, we take a look at why the subject of increasing autonomy in robots in a military context is of high significance. 

Autonomous robots – potential to revolutionise warfare

In addition to developments in the hardware of robotics with more powerful components such as sensors and processors, it is above all the methods of machine learning which enable dramatic progress to be made today. Robots are becoming more autonomous as a result of this “artificial intelligence” – both in terms of fulfilling their task in the environment, as well in completing their missions and interacting with human beings (see article «Really autonomous – or perhaps just automatic?»).

The SDRC is observing the rapid development of autonomous systems directly in practice. The two research demonstrators Armano and ANYmal have thus made enormous progress over the last three years, for example (see videos below). The increased autonomy of these two systems should benefit the Swiss Armed Forces in disaster relief. The SDRC does not pursue any research projects on the benefits of autonomous weapon systems. However, other countries are active in this area. 

Back in 2009, the American political scientist P.W. Singer explored the beginning of robot-based warfare in his book «Wired for War: The Robotics Revolution and Conflict in the 21st Century». Since then, autonomous weapons have continually been described as the third revolution in warfare (after gunpowder and atomic weapons). Autonomous weapon systems enable military superiority to be potentially ensured. With mechanical support, the loop of target search, target detection, decision on target engagement and target engagement is massively accelerated in combat operations. According to the OODA Loop Paradigm (see figure) the player who passes through this loop faster than the opponent wins the battle.

Cycle OODA-Loop

The rapid passage through this loop makes the Israeli defence system Iron Dome, for example, correspondingly valuable from a military perspective. Under constant surveillance of the airspace (observe) the system recognises approaching missiles within fractions of a second (orient), calculates their potential danger and decides whether they should be shot down (decide); if the answer is yes, the system shoots the missiles (act). A human being has no chance of performing this task in the same time period and thus in time before an event has occurred. 

Another key aspect as regards the revolution of warfare by autonomous systems is that the technologies and hardware have become more affordable and available through the massive commercial investments. The KARGU-2 drone, described in the last article in this Insights series is an example of this. The war around Nagorno-Karabakh between Armenia and Azerbaijan in autumn 2020, in which small drones were deployed, recently showed the world that more autonomous systems can indeed change warfare today. The impact of these resources made military officials around the world sit up and take notice.

This leads to the third key argument regarding why such systems could revolutionise warfare: smaller, cheaper, versatile and more capable robots mean that they can operate more and more in groups or even in swarms. And this with the potential of overwhelming the opponent’s defence. 

More autonomous robots – military advantages

In addition to accelerated operations, the reduction in costs and the scalability into groups, more autonomous weapon systems (AWS) offer numerous other military advantages: 

AWS serve to protect and relieve own forces by reducing the exposure of own soldiers in war zones. They enable a certain compensation of inventory reduction, as assigned tasks can be performed more efficiently, more permanently and more effectively with fewer staff at the front and with the corresponding technical aids – even if maintenance might be more costly. 

In addition, AWS can be used in environments in which the communication is faulty or impaired (such as faulty radio or GPS communications; as well as reduction in the electronic reconnaissance capabilities of the opponent). Operations that are tedious, difficult to endure, dangerous and time-consuming for humans (such as site surveillance, support in contaminated areas, detection and elimination of booby traps, fire fighting, etc.) will become more easily possible. 

Operations at locations that are difficult to reach or inaccessible for human beings will become easier or generally even possible through AWS (such as mountain regions, space, deep sea, etc.). 

AWS provides the opportunity to comply better with humanitarian and human rights law, as unmanned or even autonomous weapon systems do not act based on emotions (such as tiredness, stress, fear, hate, etc.). In addition, this results in the potential to reduce or even to avoid collateral damage through selective and precise impact.

More autonomous robots – military disadvantages

Some of the advantages mentioned above can become corresponding disadvantages if an opponent can benefit from them, but not one’s own armed forces. The systems – and in particular mini drones – have become more affordable and available. A weaker opponent, and not state actors, is thus also offered the possibility of being able to achieve substantial impact at a distance. This leads to a considerable effort in defending such systems, as they are being used more and more frequently. 

AWS entail the difficulty of having to distinguish between manned, unmanned and autonomous weapon systems. In addition, it is difficult to combat «swarms» of different systems (purely autonomous and intelligent or a mixture, such as drones, decoys and cruise missiles to overwhelm the integrated air defence). In addition, a further challenge is distinguishing between enemy and civilian systems, or even own systems. This is associated with an increased need for protection at the rear.

One challenge posed by AWS is to use the increased autonomy while simultaneously retaining human control over important military decisions. Where in the past, for example, a human being used to identify and register an object as hostile, it might be that today, the machine assumes this task. But should the machine now make the decision on target neutralisation? One of the main precepts of human dignity states: A human being may not be objectified. If a machine now reduces a human being to a target registered automatically with sensors, is a human being then reduced to an object? Which regulations need to be provided in order to be able to exclude this?

LAWS – positioning Switzerland and armasuisse S+T article

In order to address the various challenges and open questions on the topic of LAWS, work was begun within the framework of the United Nations in 2014. Three years later, an international group of governmental experts (GGE) for LAWS was formally created. Its mandate is to evaluate any open and arising issues in conjunction with emerging technologies in the area of LAWS. Switzerland is also participating in this discussion under the management of the Federal Department of Foreign Affairs (FDFA). The DDPS, including armasuisse S+T, supports this.

The most recent examples of the increasing significance of the topic are relevant political initiatives of the Security Policy Committee of the National Council. For example, two postulates from the Federal Council demand an examination and a report on clear regulations for autonomous weapons and the impact of drone technology on the security of Switzerland. 

For this purpose, armasuisse S+T can access its existing technical expertise and its research network. It delivers appropriate input on the required papers and discussions for the benefit of our partners in the Armed Forces and the FDFA. The SDRC manages various research projects which have corresponding points of contact. Apart from the determination of the degree of autonomy of systems (see article «Really autonomous – or perhaps just automatic?») the SDRC is also active in responding to ethical questions surrounding LAWS. For this purpose, an ethic evaluation scheme is being developed in cooperation with the universities of Zurich and St. Gallen, for example, to be able to address and make an informed assessment regarding ethical risks. Another project is investigating how consultation by a human being or artificial intelligence has an impact on the decision maker in a command support scenario. Interestingly enough, initial results have shown that those questioned actually attribute responsibility to the machine. One of the next steps of the SDRC is to systematically monitor the development of armed, more autonomous systems and to work through case studies. 

Conclusion

The systems are rapidly gaining autonomy in practice, leading to maximum, numerous military advantages. Prohibiting LAWS thus appears difficult and a profound examination of the topic seems appropriate. Just as the «Director of Studies at the Center for a New American Security» Paul Scharre wrote: «The Winner of the robotics revolution will not be who develops this technology first or even who has the best technology, but who figures out how best to use it».