Indian Defence Review

X-47B UCAS

Artificial Intelligence (AI) is the most important asset for autonomous systems. India needs to master AI technologies and applications; no one parts with them. India has to make a serious beginning to develop AI-based weapons systems and platforms to stem the excessive technological gap. DRDO has to get its act right. The theoretical research needs to be converted into formidable deliverable end-products if India is to achieve its aspirations of becoming a global player. The government needs to take the bull by the horns, allot adequate funds and position dynamic result-oriented professionals for ‘Mission AI’.

The conflict between Armenia and Azerbaijan in 2020 over the disputed Nagorno-Karabakh region saw the extensive use of missiles, drones and rocket artillery. Clearly, Azerbaijani drones were the centre of attention in this war and helped Azerbaijan take control of the skies. Azerbaijani drones provided significant advantages in ISR as well as long-range strike capabilities. They enabled Azerbaijani forces to find, fix, track and destroy targets with precise strikes far beyond the front lines. Earlier, on September 14, 2019, drones were used to attack the Saudi Aramco oil processing facilities at Abqaiq and Khurais. The very accurate attack caused precise hits and large fires at the processing facility. Both facilities had to be shut down for repairs. Meanwhile, the United States (US) has increased its use of drone strikes against targets as part of the ‘War on Terror’. On January 03, 2020, a US drone strike near Baghdad International Airport targeted and killed Iranian Maj Gen Qasem Soleimani, the commander of the dreaded Iranian Quds Force.

The American and earlier Israeli successes resulted in China, Iran, Italy, India, Pakistan, Russia and Turkey acquiring or aspiring for similar capability. An Unmanned Combat Aerial Vehicle (UCAV), also known as a combat drone, is basically a normal UAV that is capable of carrying ordnance such as missiles and bombs. These drones are usually under real-time human control, with varying levels of autonomy. The UCAVs attack targets using ‘stand-off’ weaponry, greatly depersonalising the decision to attack and considerably reduced casualties among the attackers, raising ethical questions. The advent of Artificial Intelligence (AI) resulted in greater flight and decision-making autonomy and made UAV operations free of human involvement/interference. Such UAVs possibly react more quickly and without bias, but lack human sensibility. But airborne Lethal Autonomous Robots (LARs) under a cyber-attack could go haywire and have operational and ethical implications.

Autonomous Drone Concept

The concept of ‘autonomous drones’ is that these can act based on their own choice of options or ‘system initiative’ and with ‘full autonomy‘. Such drones are programmed with a large number of alternative responses to the different challenges they may meet in performing their mission. One of the greatest challenges for the development and approval of aircraft with such technology is that it is extremely difficult to develop satisfactory validation systems, which would ensure that the technology is safe and acts like humans. At another level, autonomy could mean ‘artificial intelligence’ systems that learn and even self-develop possible courses of action.

Autonomous Aerial Combat Platforms Evolve

In 1973, Defence Advanced Research Projects Agency (DARPA) built two prototype UAVs. In the 1973 Yom Kippur war, Israel used unarmed US Ryan Firebee target-drones to spur Egypt into firing its entire arsenal of anti-aircraft missiles. Later, Israel developed the lighter Scout and the Pioneer UAVs and soon became a lead manufacturer of UAVs for real-time surveillance, Electronic Warfare (EW) and decoys. In the 1982 Lebanon War, extensive UAV-based EW resulted in no Israeli pilot being downed. The first ‘UAV war’ was the first Gulf War (May 1991) when at least one UAV was airborne at all times during Operation ‘Desert Storm’. The first human ‘kill’ by an American UAV was on October 07, 2001, in Kandahar.

Autonomous Aerial Attack Systems

Autonomous drones are systems programmed with algorithms for countless human-defined courses of action to meet emerging challenges. Onboard sensors now allow UCAVs to sense their surroundings and react accordingly, harnessing data in real time to make informed, intelligent decisions based on pre-set criteria set by the human operator. “Swarms of drones” (drones which follow and take tasks from other drones) are entirely dependent on autonomous processing. Autonomous drones that operate with manned aircraft, as unmanned “Loyal Wingman” aircraft have all been tested. Also tested are Broad Area Maritime Surveillance (BAMS) system of Poseidon P-8 maritime patrol aircraft and unmanned TRITON aircraft. Further development of UCAVs is being launched from manned aircraft, to work independently or in extension of the ‘mother aircraft’. Manned aircraft could be at the centre of a local combat or intelligence system extended with drones serving a number of supportive roles such as jamming, weapons-delivery or as multi-sensor platforms.

Artificial Intelligence Key to UAV Autonomy

An AI arms race has been on for last four-five years. The US is clear that rapid advances in AI will define the next generation of warfare. The global private investment in AI was around $70 billion in 2019. AI is a key growth investment area for the Department of Defence (DoD) of the US, with nearly $1 billion allocated in the 2020 budget. The budget for AI-supported systems is much greater. Russia has been working on AI-guided missiles that can decide to switch targets mid-flight. China is fast catching up with and plans to overtake the US in AI. The close ties between the Silicon Valley and China and the open nature of the American research community, has made the most advanced AI technology of the West easily available to China. Beijing’s roadmap aims to create a $150 billion AI industry by 2030. Beijing has committed $2 billion to an AI development park. Annual private Chinese investment in AI is around $7 billion a year. AI start-ups in China received nearly half of the total global investment in AI start-ups in 2017. The Chinese filed for nearly five times as many AI patents as did Americans. It is predicted that China will be the leading country in AI by 2025. Israel’s Harpy, anti-radar “fire-and-forget” drone, also with India, can autonomously fly over an area to find and destroy radar that fits pre-determined criteria.

Major Autonomous UCAV Systems

While the US and Israel remain world leaders in developing high technology UCAVs and autonomous platforms, China has become the ‘Wal-Mart’ of small hand-held UAVs used by hobbyists and has, of late, made significant R&D investments in military autonomous platforms. Chinese UCAV WZ-2000 is the combat version of the Xianglong High Altitude Long Endurance UAV. China is also developing a stealth strike UCAV called the ‘Warrior Eagle’ with forward swept wings, similar niche to US X-45. The future of aerial combat is here with the US military’s successful test flight of the unmanned F-16s. As per the US, drones will be used to create the most realistic ‘dog-fighting’ pilot training exercises.

BAE’s Taranis was a British technology demonstrator UCAV programme. It was part of the UK’s Strategic Unmanned Air Vehicle Experimental (SUAVE) programme with fully integrated autonomous systems and low observable features. It had a maximum take-off weight of about 8,000 kg and two internal weapons bays, making it one of the world’s larger UAVs. The first flight took place in August 2013. With the inclusion of ‘full autonomy’, the intention was for the platform to ‘think for itself’ for a large part of the mission. Taranis has now been merged into the proposed Anglo-French Future Combat Air System, where Taranis will be combined with the French Dassault nEUROn in a joint European UCAV. A test flight of a demonstrator is expected around 2025, and entry into service by around 2040. It was designated as a New Generation Fighter. Spain joined the programme in June 2019.

UCAS-D and Northrop Grumman X-47B are the US Navy (USN) successors to the US Air Force (USAF) and USN joint J-UCAS programme that was cancelled in 2006. Boeing is also working on the X-45N a concept demonstrator for a next generation of completely autonomous military aircraft. The UCAS-D programme is to demonstrate the feasibility of operating an unmanned vehicle from an aircraft carrier. Technology and operational procedures gained from the programme and X-47B demonstrator will be used to develop an operational unmanned carrier aircraft as part of the Unmanned Carrier-Launched Surveillance and Strike (UCLASS) programme. Northrop Grumman intends to develop the X-47B into an operational aircraft, the MQ-25 Stingray, which will enter service in the 2020s. The USAF has shifted its UCAV programme from medium-range tactical strike aircraft to long-range strategic bombers. The technology of the Long Range Strike programme is based on the Lockheed Martin Polecat demonstrator.

The MQ-25 Stingray Unmanned Carrier Aviation Air System (UCAAS), formerly the Carrier-Based Aerial-Refuelling System (CBARS), is a UCAV that has emerged from the UCLASS programme. In February 2016, after many delays and doubts about whether the UCLASS would specialise in strike or ISR roles, it was decided to produce a Super Hornet-sized, carrier-based aerial refuelling tanker with some ISR and some communications relay capabilities. The strike variant will evolve later. Three of these UCAVs could fly with an F-35 for refuelling and sensor operations. The MQ-25 could extend the Super Hornet’s combat radius. The competitors are Lockheed Martin’s Sea Ghost, Boeing’s ‘unnamed’ (based on Phantom Ray) and General Atomics’ Sea Avenger.

The Elbit Systems Hermes 450 is an Israeli medium-size multi-payload UCAV designed for tactical long endurance missions, with endurance of over 20 hours. Hermes 450 is equipped with two Hellfire missiles or other newer missiles. The Mikoyan SKAT is one of the Russian low-observable, sub-sonic tailless, UCAV with maximum take-off weight of ten tonne. It is meant to carry weapons in two ventral weapons bays large enough for missiles such as the Kh-31. It is powered by a single Kilmov RD-5000B turbofan engine, a variant of the RD-93. Indian DRDO’s AURA is an autonomous stealthy, flying-wing design UCAV, being developed for the Indian Air Force (IAF). It will be capable of releasing missiles, bombs and other PGMs. The programme is still in the project definition stage.

AI Enables Drone Swarming

UAV Swarming or swarm intelligence is a field of robotics research. Recent advances in chip technology and software for robotics, it has become feasible to design machines exhibiting complex behaviour, achieve mutual coordination and accomplish complex tasks. Aerial robots can ascend synchronously, communicate with each other in mid-air and create cross-references. Fixed formation group flights and complex group manoeuvres are possible. The swarm of drones behaves and functions somewhat like swarms occurring in nature, e.g., honeybee swarms, flying in coordination, displaying collective intelligence and each executing a small share of the collective task. Very small drones – some weighing less than five pounds – can cause a devastating effect if they are armed with weapons and flown in large numbers. Drone swarms can be both remotely operated, fly autonomously or may accompany ground vehicles and other aircraft. Even a single one getting through could potentially be lethal. Terrorists and other militants can also operate small, inexpensive drones loaded with weapons. Because of their size, these drones are difficult to see, hard to catch on radar and tough to shoot at with conventional weapons, particularly in swarms.

During the Opening Ceremony of the Winter Olympics at Pyeongchang, South Korea, a spectacular pre-recorded display by a quadcopter drone swarm comprising 1,218 drones, left spectators astounded. In January 2017, the USAF carried out trials with 103 Perdix quadcopter drones functioning as a swarm. The trial included air dropping of these drones in the battlefield from canisters carried by three F/A-18 fighter aircraft, gathering the drones in a swarm and then proceeding to engage targets in the battlefield. In 2016, China demonstrated drone swarming using 67 larger, fixed-wing, drones. Russia has reportedly been working on a concept of drone swarming wherein the Scandinavian countries have seen Russian drones flying in formation over their skies. Drone swarms are now being conceptualised as canister-launched weapons, especially the quadcopter ones, which would make them easy to pack and carry. These could be airdropped through fighter or transport aircraft or through bigger drones, over or close to target, depending on the danger level in the airspace in the target zone. The swarms could be varied in size depending on the task to be performed. In January 2015, the Indian Army staged a live demonstration of a swarm of 75 drones destroying a variety of simulated targets during the Army Day parade.

AI Drone Attack Options

Small quadcopter drones, laden with small but potent explosives, when employed as anti-personnel weapons could be carried hidden in the pocket and launched anywhere to target specific individuals or vital equipment. These drones could even identify the target individuals using facial recognition techniques. AI has made problem solving, target recognition, obstacle negotiation and path-finding much easier. Hundreds of drones over a battlefield or an airfield or even a political rally, would saturate the airspace and counter-swarm resources invariably would run short.

Cost Advantage of Small Autonomous Drones

Big drones are expensive, slow and vulnerable to being targeted. In contrast, small drones could be assembled into non-standard models and used to attack targets clandestinely. Since such models are cheap, they could be made in hundreds or thousands without much of a cost burden. Electronics such as GPS, digital cameras, laser range finders, RF data communication sets, processors, batteries, engines, motors and even pressure transducers and altitude sensors, are low-priced enough to be used to produce advanced capability cheap drone models for military missions including armed ones.

Drone/Swarm Counters Options

Drone swarms have some weaknesses and limitations too. First and foremost, their offensive could be blunted through the use of countermeasures like electronic warfare techniques, cyber-attacks, laser and microwave weapon systems, small arms fire, camouflage and concealment or pitching a counter drone swarm. In January 2018, Russia confirmed a swarm drone attack on its military base in Syria. Six of these small-size UAVs were reportedly intercepted and taken under control by the Russian EW units. The drones had satellite navigation electronics and carried professionally assembled Improvised Explosive Devices (IEDs).

The US is now deploying new radars such as the Q-53 system that can detect and identify such small objects and then initiate the kill chain using laser weapons. Lockheed Martin ‘Skunk Works’ engineers are doing research to develop and implement the technology that will detect and defeat drone swarms. A 60-kilowatt system that combines multiple fibre lasers to generate the high power weapon parallel beams. The laser weapon system can fire over and over, essentially creating an unlimited magazine of bullets. Laser beams are visible and can accurately target and destroy the threat at the speed of light. Cyber solutions to defeat drones are by using multi-spectral sensor systems to detect and then using cyber electromagnetic to either disable the drone or physically take over and divert. Lockheed Martin has already supplied the US Army with a 60-kilowatt laser weapon mounted on a large modified truck that can destroy rockets, artillery, missile, drones and other trucks or ground vehicles. These can also be integrated onto aircraft, ground vehicles and ships.

Ethical and Legal Issues

Autonomous drones, when they are used during armed conflict, would be subject to the general principles and rules of the Law of Armed Conflict. In this respect, autonomous drones are not to be distinguished from any other weapon system. The question is for how long may an autonomous weapons system (lawfully) be ‘left alone’ to operate (for hours or days)? The delegation of life-and-death decisions to non-human agents is being questioned by those who oppose autonomous weapons systems.

As with any ‘means of warfare’, autonomous drones must only be directed at lawful targets (military objectives and combatants) and attacks must not be expected to cause excessive collateral damage. Attacking humans with remote-controlled machines was already a level higher than the use of other ‘stand-off’ weaponry such as missiles, artillery and aerial bombardment, because of depersonalised decision to attack. It now gets even more complicated if the UAV can initiate an attack autonomously. Drones are more likely to be hacked if they are autonomous otherwise the human operator would be able to take control. Technological possibility of autonomy should not obscure the continuing human moral responsibilities.

Though accuracy has greatly reduced collateral damage, drones are still blamed for a considerable number of innocent civilian deaths. With war becoming safer and easier, as soldiers are removed from the horrors of war and see the enemy not as humans but as blips on a screen, there is very real danger of losing the deterrent that such horrors provide. Controllers can also experience psychological stress from the combat they are involved in. A few may even experience Post-Traumatic Stress Disorder (PTSD). Unlike bomber pilots, drone operators see its effects on human bodies in stark detail. Limiting the risk to soldiers by removing them from the battlefield altogether could make war too ‘easy’, reducing it to a low-cost technological game that no longer requires any public or moral commitment. One school of thought is that our fundamental responsibility for war and how wars are fought can never be morally ‘outsourced’, least of all to machines. The laws are still evolving. In the meantime, the world requires a reasonable Commander to act in good faith.

Autonomous Systems Way Ahead

The autonomy of current systems is restricted where the craft operates autonomously under certain conditions, but a pilot must monitor its progress. The next level will be the craft is autonomous in most situations; the pilot can take over, but generally does not have to. The final level will be when the drone is fully autonomous. To ensure airspace safety with fully autonomous drones and aircraft will require new rules, air/road traffic control systems, pre-defined routes and enhanced technology to sense, react and avoid obstacles. The ability of some vehicles to transition from hover to lift-based forward flight and vice-versa brings the possibility for an autonomous flying vehicle to perform complex missions where the two different flight modes are needed. Due to their small size, autonomous UAVs are often sensitive to environmental disturbances such as wind gust. Control laws based on the super-twisting algorithm are being evolved.

Autonomous weapons are today capable of deciding a course of action from a number of alternatives. Although the overall activity of an autonomous unmanned aircraft will be predictable, individual actions may not be. DARPA has been developing a fleet of small naval vessels capable of launching and retrieving combat drones without the need for large and expensive aircraft carriers. The Pentagon is looking at ideas on how to build a flying aircraft carrier that can launch and retrieve drones using existing military aircraft such as the B-1, B-52 or C-130. The US is developing new undersea drones that can operate in shallow waters where manned submarines cannot. The Pentagon’s ‘Blue Sky’ research project aims at developing ‘intelligent machines’. Russians have had robots armed with grenade launchers and Kalashnikovs. China too is investing heavily in automated weapons systems and platforms.

India – Getting its Act Right

The first combat capable version of the Pakistani Burraq UCAV was first publicly demonstrated in March 2015. Autonomous unmanned systems are where the future is. In view of small defence expenditures and the persisting duplications of military capacities, mixed manned and unmanned air formations might be opportunity for future conflicts. Intensive weapon research is going on for AI-based autonomous weapon systems. That is where the future is. India is part of the most threatened regions of the world and needs to monitor weapon developments closely. With very few players in the market, technologies are closely guarded.

At Aero India 2021, Hindustan Aeronautics Ltd (HAL) demonstrated the manned-unmanned-teaming system called the “Combined Air Teaming System” (CATS). Essentially, the LCA two-seater is proposed as the mother aircraft. It is meant to control two to four “Warriors” launched independently. CATS “Hunter” UAV will be carried on the LCA pylons. “Alpha” Air Launched Flexible Asset will give shape to a swarm, each carrying a five to eight kilogramme payload and with the capability to strike multiple targets. Eventually, the concept will be ported on transport aircraft and helicopters.

AI is the most important asset for autonomous systems. India needs to master AI technologies and applications as no nation would part with them. India has to make a serious beginning to develop AI-based weapons systems and platforms to stem the excessive technological gap. DRDO has to get its act right. The theoretical research needs to be converted into formidable deliverable end-products if India is to achieve its aspirations of becoming a global player. The government needs to take the bull by the horns, allot adequate funds and position dynamic result-oriented professionals for ‘Mission AI’.