AI Reshapes Air Defense Command: New Insights from Zhengzhou Campus Experts
As global military powers accelerate their race into the age of artificial intelligence (AI), a new wave of transformation is sweeping through command and control systems, particularly in air defense. At the forefront of this evolution, researchers from the Zhengzhou Campus of the Army Artillery and Air Defense Academy have published a comprehensive analysis outlining how AI can redefine the future of air defense command and control systems. The study, led by Bao Zuohui, Yang Zuobin, and Sun Danhua, was featured in Ordnance Industry Automation, a respected technical journal in China’s defense sector, offering timely insights into the technological, operational, and strategic shifts driven by intelligent warfare.
The paper, titled Development and Construction of Air Defense Command & Control System in AI Era, presents a forward-looking framework that identifies three critical pillars for next-generation systems: enhanced situational awareness, intelligent decision support, and advanced human-machine interaction. These domains are not merely incremental upgrades but represent a fundamental rethinking of how air defense operations will be conducted in an era where speed, complexity, and uncertainty dominate the battlefield.
Published in November 2021, the research comes at a pivotal moment when autonomous platforms, hypersonic threats, stealth aircraft, and swarming drones are redefining aerial combat. Traditional command structures, built on linear information flows and manual processing, struggle to keep pace with the velocity of modern engagements. As adversaries deploy increasingly agile and unpredictable tactics—evidenced by recent conflicts such as the Nagorno-Karabakh war, where drones played a decisive role—the need for faster, smarter, and more adaptive command systems has never been greater.
Bao Zuohui, a lecturer specializing in command information system applications, emphasizes that while current air defense networks have achieved integration across echelons—from corps-level down to battalion and company units—they still operate largely within deterministic frameworks. “We’ve moved from analog to digital, from isolated systems to networked architectures,” he explains. “But true intelligence lies beyond connectivity—it lies in cognition.”
This distinction forms the core of the authors’ argument: today’s systems excel at data transmission and display, but fall short in interpretation, prediction, and proactive reasoning. They function more like high-speed pipelines than thinking entities. In contrast, AI-driven systems aim to close this gap by transforming raw sensor inputs into actionable knowledge, enabling commanders to anticipate enemy moves rather than simply react to them.
One of the most compelling aspects of the study is its alignment with the broader trajectory of AI development. Drawing from established models, the authors categorize AI progress into three tiers: computational intelligence, perceptual intelligence, and cognitive intelligence. Computational intelligence refers to number-crunching capabilities—such as those used in early expert systems for fire distribution or route planning. Perceptual intelligence involves pattern recognition, exemplified by deep learning algorithms that process images, signals, or speech. Cognitive intelligence, the highest level, encompasses understanding, inference, and decision-making under uncertainty.
According to the team, U.S. military programs such as “Deep Green,” “Insight,” and the “Commander’s Virtual Staff” have already begun exploring cognitive-level applications. For instance, Deep Green aimed to predict battlefield outcomes seconds or minutes ahead using simulation-based forecasting—a capability that could dramatically compress the OODA loop (Observe, Orient, Decide, Act). Meanwhile, China’s efforts, though initially focused on rule-based expert systems, are now shifting toward machine learning and data-driven modeling.
The Chinese researchers acknowledge that domestic advancements started later, particularly after the symbolic defeat of chess champion Garry Kasparov by IBM’s Deep Blue in 1997, which served as a wake-up call for many nations about the disruptive potential of AI. Since then, significant investments have flowed into AI research, especially in areas relevant to defense. However, the transition from automation to genuine autonomy remains incomplete.
To bridge this gap, the authors propose a tripartite strategy centered on what they describe as the “three intelligences”: intelligent battlefield situation awareness, intelligent decision recommendation optimization, and intelligent human-computer collaborative interaction. Each component addresses a distinct bottleneck in current command workflows.
Reimagining Situational Awareness: From Data Flood to Strategic Insight
Modern battlefields generate vast amounts of heterogeneous data—radar returns, electro-optical feeds, electronic intelligence, satellite imagery, and communications intercepts. Individually, these sources provide fragments; collectively, they form a chaotic mosaic. Current fusion techniques often rely on pre-defined correlation rules, which work well under stable conditions but falter when faced with novel scenarios or adversarial deception.
The solution proposed by Bao, Yang, and Sun hinges on integrating knowledge-based reasoning with big data analytics and predictive inference engines. Their vision goes beyond mere data aggregation; it aims to create a dynamic, self-updating understanding of the operational environment.
At the foundation is knowledge-base-driven situational fusion. Rather than treating all incoming reports equally, the system uses a structured knowledge repository—initially populated with expert-derived rules and doctrines—to interpret events in context. Over time, through Bayesian network learning applied to historical engagement data, the knowledge base evolves, refining its accuracy and adaptability. This allows the system to distinguish between routine movements and potentially threatening patterns, even when data is sparse or contradictory.
Complementing this is big-data-enabled situational analysis. With petabytes of real-time sensor output streaming in, traditional databases cannot keep up. Instead, the authors advocate for distributed computing frameworks capable of handling the “4V” characteristics of military big data: volume, variety, velocity, and veracity. By applying data mining, feature extraction, and anomaly detection algorithms, the system can identify hidden correlations—such as the link between radar silence and missile launch preparation—or detect subtle indicators of coordinated attacks across multiple domains.
Perhaps most transformative is the concept of inference-engine-powered situational prediction. Here, deep learning models trained on thousands of simulated and real-world engagements learn to forecast adversary behavior based on evolving battlefield conditions. For example, if enemy UAVs begin loitering near a protected zone while jamming intensifies, the system might infer an imminent strike and recommend preemptive countermeasures. Such predictions are not guesses but probabilistic assessments grounded in doctrinal norms, equipment performance limits, and environmental constraints.
Together, these technologies enable a shift from reactive monitoring to anticipatory cognition—a leap akin to moving from weather observation to meteorological forecasting.
From Assistance to Anticipation: Rethinking Decision Support
While situational awareness provides the “what” and “where,” decision support tackles the “so what” and “what next.” Historically, command systems offered tools—maps, calculators, templates—but left the actual planning to human operators. Even today, mission planning modules require extensive manual input, limiting responsiveness during fast-moving crises.
The authors argue that AI must evolve from passive assistants to active collaborators. To achieve this, they outline three complementary approaches.
First, task-planning-based intelligent decision recommendations leverage AI search algorithms and knowledge graphs to automate complex planning tasks. Tasks such as force deployment, reconnaissance coverage, coordination timelines, and resource allocation involve combinatorial complexity far beyond unaided human capacity. By encoding tactical principles and operational constraints into formal logic, AI planners can generate multiple viable options in seconds, ranked by feasibility, risk, and expected outcome.
Second, autonomous-decision-based scenario simulation and optimization introduces adaptive agents into wargaming environments. Unlike static simulations that follow fixed scripts, these models allow individual units—represented as intelligent nodes—to make independent decisions based on changing circumstances. During a rehearsal, if a simulated enemy changes course, friendly units dynamically adjust their responses, revealing vulnerabilities or opportunities invisible in rigid walkthroughs. Post-simulation evaluation metrics help refine plans iteratively, leading to more robust and flexible strategies.
Third, machine-learning-enhanced ad hoc decision support addresses the chaos of real-time emergencies. When unexpected events occur—such as sudden radar failure or an unanticipated incursion—commanders often rely on intuition and experience. The proposed system enhances this by rapidly scanning a database of historical cases, doctrine references, and contingency plans, then applying natural language processing and similarity matching to suggest optimal responses. It doesn’t replace judgment; it accelerates it.
Crucially, the researchers stress that full autonomy is neither desirable nor feasible in high-stakes military contexts. Human oversight remains essential. But by offloading routine cognitive labor, AI frees commanders to focus on higher-order thinking—strategy, ethics, escalation management.
Human-Machine Symbiosis: Redefining Interaction Design
Even the most sophisticated AI is useless if users cannot interact with it efficiently. Poor interface design creates friction, delays, and errors—especially under stress. Yet many existing command systems still depend on legacy paradigms: keyboard-and-mouse navigation, cluttered menus, and fragmented displays.
Recognizing this, the third pillar of the authors’ framework focuses on intelligent human-computer collaboration. Inspired by multimodal interaction concepts seen in consumer robotics and virtual assistants, they envision a future where operators engage with command systems through voice, gaze, gesture, and even neural signals.
Imagine a commander issuing orders verbally—”Engage target Bravo with Battery 3″—while eye-tracking confirms intent by highlighting the correct icon on screen. Or consider a scenario where muscle signals from a wearable sleeve trigger rapid-fire commands without touching any device. Such interfaces reduce physical workload, minimize distraction, and enhance speed.
Moreover, the team advocates for context-aware interaction prioritization. Not all actions carry equal urgency. Launching surface-to-air missiles requires different handling than updating logistics status. Therefore, the system should adapt its response mode based on task criticality. Time-sensitive directives could be executed via dedicated hardware buttons or voice shortcuts, ensuring one-step activation. Less urgent functions remain accessible through standard menus.
Security and reliability are paramount. Multimodal verification—requiring both voice command and retinal scan, for instance—can prevent unauthorized access or accidental execution. Additionally, explainable AI components ensure that every automated suggestion includes transparent rationale, preserving trust and accountability.
These enhancements go beyond convenience; they reshape the ergonomics of command. In high-tempo operations, shaving seconds off reaction time can mean the difference between interception and penetration.
Strategic Implications and Future Trajectory
The implications of this research extend beyond technical innovation. It reflects a deeper shift in military philosophy—one that embraces AI not just as a tool, but as a co-evolving partner in warfare. As Bao notes, “We’re no longer building systems to execute commands. We’re building systems that help us understand the battlefield before we even issue them.”
This paradigm aligns with global trends. The United States’ Third Offset Strategy, initiated in the 2010s, explicitly sought technological superiority through AI, autonomy, and cyber capabilities. Similarly, Russia and other major powers have unveiled ambitious roadmaps for robotic combat units and AI-integrated C4ISR (Command, Control, Communications, Computers, Intelligence, Surveillance, and Reconnaissance) networks.
China, too, has elevated AI to national strategic priority, exemplified by President Xi Jinping’s emphasis on intelligentized warfare and the release of national AI development guidelines. Within this context, the work of Bao, Yang, and Sun represents both a contribution to academic discourse and a practical blueprint for institutional modernization.
However, challenges remain. Data quality, algorithmic bias, adversarial attacks on AI models, and ethical concerns around lethal autonomy require careful governance. Moreover, organizational resistance, interoperability issues, and talent shortages may slow adoption.
Yet the momentum is undeniable. As AI transitions from narrow applications to integrated cognitive ecosystems, air defense command systems will become less centralized hubs and more distributed nervous systems—adaptive, resilient, and continuously learning.
In their concluding remarks, the authors quote a powerful metaphor from earlier literature: “Artificial intelligence is like a train. You hear it coming, you wait for it, it arrives—and then it speeds past, leaving you behind.” The message is clear: hesitation is not an option. Those who fail to prepare for the AI revolution in command and control risk being overwhelmed by adversaries who do.
For defense technologists, policymakers, and military leaders worldwide, the study serves as both a warning and a roadmap. The future of air defense is not just automated—it is intelligent. And the journey has already begun.
Bao Zuohui, Yang Zuobin, Sun Danhua, Zhengzhou Campus of Army Artillery and Air Defense Academy, Ordnance Industry Automation, doi: 10.7690/bgzdh.2021.11.005