FUTURE ARTIFICIAL INTELLIGENCE BASED AUTONOMOUS WAR (PART 1)…

The character of war is changing. Concepts and ideas such as grey-zone warfare[i], hybrid warfare[ii], fifth generation warfare[iii] and the likes are being used extensively in an ever-evolving technological environment and this is possible due to the miniaturization of future carriers and also of the delivery system. In future we expect to see miniature weapons delivery systems which are likely to be semi or fully autonomous. Land, maritime and air-based platforms are available with the latter having much more flexibility in terms of reach, deployability and lethality.

The once considered ‘King of Battlefield’ — Tank, was reduced to a mere ‘Pawn’ by the ‘killer’ Unmanned Combat Aerial Vehicle (UCAV) Bayraktar TB2[iv] supplied by Turkiye to Azerbaijan during the Nagorno-Karabakh conflict with Armenia.

The Bayraktar TB2 a semi-autonomous weapon system acted as a game changer and contributed significantly towards the victory of Azerbaijan over Armenia. In fact, it made it a one-sided war as Armenia had no counter to the accuracy of attacks by Bayraktar UCAV and the Armenian’s armoured columns appeared as ‘soft targets.’ In effect, Turkiye showed its expertise in the field of UCAV and marked a significant improvement of capabilities of Bayraktar than ever before. Besides, it made the world sit-up and notice the role of ‘Autonomous War System (AWS)[v]. In addition to this, Azerbaijan, also displayed the Israeli made ‘Harop’ that could be fired from a mother aircraft, loiter over the target area for hours to identify, select and attack the defined target.

The Artificial Intelligence based systems which were hitherto fore being used in civilian domain for operating machines and home gadgets are now increasingly being used in war for planning and execution and in weapon systems. Artificial intelligence and automation are indeed the future of war. In the last three years, some countries have shown considerable advancement towards this novel idea for waging wars that is smarter, faster, efficient and precise.

In the battlefields of the future, instead of soldiers, mother vehicles/crafts might simply offload fleets of AI-controlled tanks and other armed vehicles and drones. Those tanks could be equipped with systems that would seek and train their arsenal on targets up to three times faster than humans. This system automatically detects, evaluates, identifies, tracks, selects, engages, and conducts battle-damage assessments. It can take out any target it is aimed at. All of this without human intervention and only using AI based data, is indeed what the future of war would look like[vi].

Concept of War Fighting

The four steps of decision making viz; Observe, Orient, Decide and Act (OODA) is a recurring cycle in war fighting. This cycle commences when the war starts and continues till it terminates. It functions on a continuous feedback mechanism. After every action, feedback is taken before the next step is initiated. It prevents resource and time wastage and improves efficiency of war waging potential. Between two opponents, the one who is able to have a faster OODA loop will get the upper hand leading to victory.

• Observe refers to the raw information received as a result of surveillance carried out by multiple resources (both human and electronic). The information needs to be processed, filtered and disseminated in the right format to be usable.

• Orient refers to the individual or team which has been provided the intelligence (filtered observation), appropriately gets ready to face the opponent and also carry out threat analysis based on training and experience. It is about getting aware of the evolving situation.

• Decide on the course of action / option to be taken from among the many and weighing their potential outcome before choosing one. It is about making the right choice based on the many parameters which are presented, even in the face of the enemy.

- Act on the decision made. Once the action is initiated and the enemy reacts, the next cycle of OODA loop commences.

Impact of AI on OODA loop

The importance of Network Centric Warfare (NCW)[vii] for future wars has been realised, but NCW in itself may not be sufficient to dominate information and control its flow. As stated, the OODA loop is based on the fundamental premise that the loop is faster than the opponent side and can contribute to success. To reduce OODA loop, the decision-making authority has to act fast. Technological advancement has resulted in voluminous data flow from multiple sources, to be analysed in considerably shorter time frame. To analyse the whole of it and filter out the useful information remains a struggle. Data Analytics and AI can enable armed forces to scan data and arrive at insights that can help them gain an edge. The OODA loop puts big data and analytics into a decision-oriented frame of reference that allows a leader to effectively take advantage of technologies. Al would assist the authority to complete OODA cycle in shorter time and to take decision quickly.

• Observe. If the general area of operations is known, a large quantum of surveillance data can be pre-stored. Similarly, based on enemy’s and own likely pattern of operations and equipment being used, all known data can be pre-fed and managed. The data management involves inventorying data sources, identifying the useful data fields, building a data catalogue, etc. When actual operation commences, using AI based data correlation, useful information / intelligence will be easily obtained in a quicker time frame. This would allow a leader to accurately focus on resources and make better decisions.

• Orient. All leaders have varied perceptions through which they understand their own and opponent’s environment. At the orientation phase, leaders apply context to the data collected, and this in turn helps them create situational awareness. AI helps in sorting and prioritizing the data so that it can be organized and recalled, in a ready to use format, based on the leader’s perception of impending operation.

AI handles several functions to include:

• Operation-stream processing[viii], which continuously queries data in motion to detect patterns and analyse events in real time.

• In-memory analytics[ix], which allows complex data exploration, selected deployment to be processed in memory and distributed in parallel across a dedicated set of nodes.

• Visual analytics[x], which leverages in-memory technology to identify patterns and relationships in data that were not obvious before.

• Visual statistical data[xi], which combines predictive analytics with visual data exploration capabilities in a single, interactive environment to run options more quickly.

• Decide. AI can automate cognitive thinking to change the way a leader perceives information and makes decisions, rather than being programmed to anticipate every possible action needed to perform a function or set of tasks, humans can sense, predict and infer. These analytics solutions can also process natural language and unstructured data and learn by experience to help leaders make better decisions based on the best options available.

- Act. After developing and accessing multiple options, the decision is put into action. All the data collated is worthless if it cannot be used to accelerate effective action. Big data and analytics can help compress the decision cycle. By enlisting the OODA loop, leader can put themselves in an offensive rather than a defensive position against their opponents. Thereafter, the next OODA cycle is repeated.

[i] Grey-zone activities are coercive statecraft actions short of war. The grey-zone is a mainly non-military domain of human activity in which states use national resources to deliberately coerce other states.

[ii] Hybrid Warfare entails an interplay or fusion of conventional as well as unconventional instruments of power and tools of subversion. These instruments or tools are blended in a synchronised manner to exploit the vulnerabilities of an opponent.

[iii] Fifth Generation Warfare is warfare that is conducted primarily through non-kinetic, such as social engineering, misinformation, cyber-attacks, along with emerging technologies such as artificial intelligence and fully autonomous systems.

[iv] Bayraktar TB2 Armed Unmanned Aerial Vehicle available at https://www.turkishdefencenews.com/bayraktar-tb2-armed-unmanned-aerial-vehicle. Assessed on 28 Jun 24.

[v] Amitai Etzioni and Oren Etzioni, “Pros and cons of Autonomous Weapons Systems”, Military Review dated May-June 2017. Available at https://www.armyupress.army.mil/Journals/Military-Review/English-Edition-Archives/May-June-2017/Pros-and-Cons-of-Autonomous-Weapons-Systems. Assessed on 28 Jun 24.

[vi] Loukia Papadopoulos, “The Scary, Automated Future of War”, Interesting Engineering, 17 Dec 21 available at https://interestingengineering.com/video/the-scary-automated-future-of-war . Assessed on 28 Jun 24.

[vii] Raymond McConoly, “What is Network Centric Warfare?” Naval Post, 21 June 21 available at https://navalpost.com/what-is-network-centric-warfare. Assessed on 28 Jun 24.

[viii] Harsh Varshney, “What is Stream Processing?” HEVO 8 April 22 available at https://hevodata.com/learn/stream-processing/. Assessed on 05 Jun 22.

[ix] In-memory analytics is an approach to querying data when it resides in a computer’s random-access memory (RAM), as opposed to querying data that is stored on physical disks.

[x] Visual analytics is the use of sophisticated tools and processes to analyze datasets using visual representations of the data.

[xi] Visual statistics brings the most complex and advanced statistical methods within reach of those with little statistical training by using animated graphics of the data.