Elsa Kania is an Adjunct Fellow with the Technology and National Security Program at the Center for a New American Security, where her research focuses on Chinese defense innovation and emerging technologies, particularly artificial intelligence. Elsa is an analyst, a consultant, and a co-founder of the China Cyber and Intelligence Studies Institute.
As the character of conflict is transformed by the advent of robotics and artificial intelligence on the battlefield, the Chinese People’s Liberation Army (PLA) recognizes and seeks to capitalize upon this trend towards “unmanned, intangible, silent warfare” (無人, 無形, 無聲戰爭) that is increasingly “intelligentized” (智能化). Consequently, the PLA has prioritized advances in military robotics and ‘unmanned’ (i.e., uninhabited) systems. To date, the PLA has fielded a range of unmanned aerial vehicles (UAVs), while also developing and, to a limited extent, fielding unmanned underwater vehicles (UUVs), unmanned ground vehicles (UGVs), and unmanned surface vehicles (USVs). Concurrently, the Chinese defense industry is actively pursuing research and development for a range of cutting-edge unmanned systems, including those with stealth, supersonic, and swarming capabilities. The PLA is also prioritizing the development of military applications of artificial intelligence (AI), including to enable data and intelligence fusion and to support command decision-making. In the near future, unmanned and autonomous systems could serve as a force multiplier for the PLA’s combat power.
For Taiwan, the PLA’s pursuit of and ongoing advances in these capabilities are cause for concern. As Ian Easton of the Project 2049 Institute noted in his recent book, The Chinese Invasion Threat, such a scenario would likely include the use of unmanned systems to support initial strikes and an amphibious assault alike. The PLA appears to be preparing to leverage unmanned systems for a range of missions, including intelligence, surveillance, and reconnaissance (ISR); integrated reconnaissance and strike; information operations, especially electronic warfare; and data relay, including communications relay and guidance for over-the-horizon (OTH) targeting. Notably, continued progress in swarm intelligence (集群智能) could enable asymmetric assaults against major US weapons platforms, such as aircraft carriers. For instance, China’s Military Museum includes in one exhibit a depiction of a UAV swarm combat system (無人機蜂群作戰系統) with swarms used for reconnaissance, jamming, and “swarm assault” (群打擊) targeting an aircraft carrier. Recognizing the potential of “saturation attacks” to overcome even sophisticated defenses, the PLA could leverage similar tactics against Taiwan, and reportedly there have been efforts to convert retired fighter jets for this purpose. Concurrently, the PLA Navy is likely to acquire and employ unmanned surface vessels (USVs) that could be used troop transport or logistic support, and the PLA Marine Corps might even utilize unmanned tanks or amphibious combat vehicles in support of a future landing campaign.
The PLA Army, Navy, Air Force, Rocket Force, and Strategic Support Force all operate several different models of UAVs. For instance, the PLA Army uses the ASN-207 for such tasks as battlefield reconnaissance, communications relay, and electronic warfare. Certain PLA ground forces, likely including special forces, operate a smaller, hand-held and launched variant, the CH-802, that could be used to enable situational awareness on the battlefield at the tactical level. The PLAN operates the medium-altitude long-endurance (MALE) BZK-005 for surveillance in the East and South China Seas, including in proximity to Taiwan, while the medium-altitude, medium-endurance (MAME) UAV, the ASN-209, can be used for communications relay and electromagnetic countermeasures. The PLAAF has fielded the GJ-1 (Gongji-1, 攻擊－1), a MALE UAV roughly analogous to the US Predator, for use in integrated reconnaissance and precision strike, and may soon introduce the GJ-2, a successor that is closer to the Reaper in capabilities. The PLA Rocket Force has fielded at least a limited number of UAVs, including to Base 52 in Anhui Province, potentially to provide over-the-horizon guidance for missiles, such as the DF-21D. The PLA’s new Strategic Support Force (SSF) and/or the Joint Staff Department’s Network-Electronic Countermeasures Dadui (網電對抗大隊) could also field UAVs in support of their electronic warfare missions. Of note, the military parade that celebrated the PLA’s 90th anniversary last August also featured the ASN-301 anti-radiation loitering munition system, characterized as a “radar killer,” which was developed based on a reverse-engineering of Israel’s Harpy drone to target enemy radio and communications stations. The PLA is continuing to develop and preparing to field new types of UAVs, including the Xianglong (“Soar Dragon,” 翔龍) a high-altitude long endurance (HALE) UAV, for missions of long-range reconnaissance and/or electronic warfare.
Looking forward, it is clear that in any crisis or conflict scenario, Taiwan could be confronting a panoply of PLA unmanned systems across all domains of warfare. The ongoing progress of the PLA’s research, development, testing, and fielding of these “new-type forces” (新型力量) will merit continued analytical attention. Certain PLA thinkers even anticipate that intelligent machines become primary warfighters in future militaries, while humans remain planners, administrators, and commanders. However, if the PLA progresses towards greater reliance upon highly automated and autonomous systems, this change in the character of warfare could also create new challenges and perhaps unexpected vulnerabilities. Despite its notable progress in research and development, the PLA could confront continued difficulties in human talent and training, given the complex dynamics of human factors associated with such systems. For instance, the PLA Air Force’s introduction of the GJ-1 has required major efforts in talent cultivation. In an actual conflict scenario, unmanned systems could also have limited survivability on the battlefield, due to potential vulnerability to interference with their data links and control mechanisms. Although the PLA does train for the use of unmanned systems in complex electromagnetic environments, their actual level of resilience against electronic countermeasures in a conflict scenario remain to be seen. It will be critical for Taiwan to focus on enhancing its electronic warfare capabilities as the PLA advances in complex weapons systems that are dependent upon battle networks.
At present, the creation of counter-drone and counter-swarm capabilities remains an open challenge in which creativity and targeted investments might enable Taiwan to achieve an asymmetric advantage. To date, the use of directed-energy weapons, such as lasers or high-power microwave weapons, could have utility in targeting drones. There have also been tests of lower-tech alternatives, such as the use of a net. Taiwan should consider actively exploring options and investments. Even if the PLA starts to leverage AI for military purposes and introduces further systems with higher degrees of autonomy, these new capabilities could cause concurrent vulnerabilities. For instance, the use of computer vision (e.g., to enable automated processing of imagery and video, or for automatic target recognition) is often characterized as a more mature application, which the United States and China, among others, are pursuing. However, there have also been clear demonstrations of the extreme and even unexplainable exploitation of the underlying neural networks based adversarial examples, even as seemingly minor as changes to a single pixel. These and other potential algorithmic vulnerabilities might be exploited and weaponized. As the PLA progresses in its efforts to take advantage of AI systems, it could become ever more vulnerable to future “counter-AI” capabilities.
Today’s rapid advances in AI and robotics could catalyze a new military revolution. Lieutenant General Liu Guozhi (劉國治), director of the Science and Technology Commission at the Central Military Commission believes that AI will accelerate a process of military transformation. He warns, “facing disruptive technology, [we] must … seize the opportunity to ‘change paradigms’ (彎道超車); if you don’t disrupt, you’ll be disrupted!” Although China is already a major force in robotics and aspires to “lead the world” in AI, the PLA’s capability to engage in successful defense innovation will, however, depend upon human factors and its organizational capability. In these domains, chronic issues in the PLA’s strategic, command, and organizational cultures could exacerbate the difficulty of making the deeper changes that would be necessary to leverage these new technologies. While seeking to achieve a disruptive advantage, the PLA also faces continual risks of disruption. In this context, Taiwan could build upon its own progress in unmanned systems and its new national initiative and advances in AI to pursue the development of asymmetric capabilities that leverage these technologies, while exploring creative means of countering them.
The main point: In any crisis or conflict scenario, Taiwan could confront a panoply of PLA unmanned, even autonomous, systems across all domains of warfare. Although the PLA’s pursuit of these disruptive new capabilities could thus intensify the threats that Taiwan is facing, there are also opportunities for Taiwan to seek disruption in response, including through advancing its electronic warfare capabilities and exploring options for counter-drone, counter-swarm, and counter-AI capabilities.
¹軍事科學院軍事戰略研究部 [Academy of Military Science Military Strategy Research Department], 戰略學[The Science of Military Strategy], 軍事科學出版社 [Military Science Press], 2013, p. 97-98.
²See, for instance: Liang Yong [梁勇] and Zhou Shaolei [周紹磊], “UAV Over-the-Horizon Guidance Methods” [無人機超視距引導方法], Missile and Aerospace Delivery Technologies [導彈與航天運載技術], 2010. The authors are affiliated with the Naval Aeronautical Engineering Institute’s Control Engineering Department.
³Although the literal translation of this term implies “cutting past the car in front of you around a corner”, I choose to use a more figurative translation that I think better conveys the intended meaning in this context (i.e., disruption that enables the PLA to surpass the US military.)