中國軍事大數據：軍事情報轉型的加速器

來源：解放軍報 作者：宋元剛、邵龍飛、特約通訊員 王涵 主編：吳行健

現代英語：

In order to thoroughly implement President Xi’s decision to “promote the implementation of the national big data strategy” and accelerate the development of military intelligence, our military has begun to take active actions. The innovative application of military big data will surely stimulate data vitality, release data value, and produce a multiplier effect like never before, so as to open the door to future victory.

In order to actively adapt to the new situation and new requirements, promote the deep integration of big data and national defense construction, and promote exchanges and cooperation among experts in the field of military big data research, the Second Military Big Data Forum, hosted by the Academy of Military Sciences and themed “Military Big Data Promotes the Development of Military Intelligence”, was held in Beijing from August 22 to 23 this year. More than 500 leaders, experts and representatives from the Central Military Commission, various theater commands, various services, the Academy of Military Sciences, the National Defense University, the National University of Defense Technology and other units participated in the discussion and exchange on the frontier, common and hot issues of the development of military big data.

During the forum, our reporter interviewed Researcher Liu Linshan, Director of the Military Science Information Research Center of the Academy of Military Sciences, Researcher Lu Bin, Deputy Director, and Associate Researcher Luo Wei, Director of a research laboratory, on topics related to military big data. 

Military big data is highly confrontational

Reporter: my country is currently vigorously promoting the implementation of the national big data strategy. Big data is being widely used in many fields such as economy, politics, security and social management, reflecting unprecedented great value. So, compared with civilian big data, what are the connotations and characteristics of military big data?

Liu Linshan: With the development of big data technology and applications, military big data has broken through the concept of military data in the past and has become a general term for a series of activities based on massive military data resources, with data intelligent processing and analysis technology as the core, and driven by the extensive application needs in the military field. In view of the particularity of military activities, in addition to the typical characteristics of civilian big data such as large data scale, multiple content types, high processing speed, and low value density, military big data also has the characteristics of “one super, one high, and one strong”.

“One super” means super complexity. It means that the data comes from multiple spaces such as land, sea, air, space, electricity, and the Internet. The information dimension is higher, the unstructured characteristics are more obvious, and the data relationship is more complex. “One high” means high security, which means that the threats faced are complex, including enemy reconnaissance and theft, leakage of confidentiality by one’s own side, system vulnerabilities, and attacks by the enemy’s “soft” and “hard” means, etc., and the risk of weakening or losing availability is greater. “One strong” means strong confrontation, which means that the game confrontation between information acquisition and anti-acquisition means, the widespread existence of data fog disguise and deception, and the intricate interweaving of true and false data, which requires extremely high ability to distinguish the authenticity of data.

Lv Bin: Here we need to emphasize the “strong confrontational” characteristics of military big data. Since military big data is data in a confrontational environment, the data quality is poor, the value density is low, and it is usually uncertain, incomplete, and false. We know that artificial intelligence at this stage is mainly based on data-driven machine learning. Machine learning requires sample data, but the current war is small sample data, and future wars may not even have sample data. In addition, the characteristic of military operations is “man in the loop”, and human activities are difficult to learn and predict using classic big data methods. This makes military big data much more difficult than civilian big data in dealing with problems such as small sample data learning, game under incomplete and uncertain information, and scene modeling and understanding in complex environments. The challenges encountered are much greater, and new theories, new methods, and new technologies must be adopted to solve them.

Military big data and military intelligence complement each other

Reporter: The report of the 19th CPC National Congress emphasized the need to “accelerate the development of military intelligence and improve joint combat capabilities and all-domain combat capabilities based on network information systems.” How should we understand the relationship between military big data and military intelligence?

Liu Linshan: Looking back at the development of artificial intelligence, since it was first proposed at the Dartmouth Conference in 1956, it has experienced three climaxes and two troughs: the reasoning period, the knowledge period, and the learning period. Scientists have tried to “manufacture” human intelligence through logical reasoning, expert systems, etc., but the failure of the development of Japanese intelligent computers and the decline of the Encyclopedia of Human Common Sense Knowledge at Stanford University in the United States have made these paths unsustainable. The emergence of big data has provided a new path for the development of artificial intelligence, and people have begun to shift from how to “manufacture” intelligence to how to “learn” intelligence. Different from the traditional hope of achieving reasoning and learning through rules, logic and knowledge, through machine learning, we can gain insight into the hidden laws of massive data from big data, and effectively realize data-driven artificial intelligence. In March 2016, the “AlphaGo” Go robot developed by Google was able to defeat the world champion Lee Sedol based on the deep learning of neural networks based on 30 million chess records. In addition, the development of big data technology can also make up for the shortcomings of artificial intelligence in algorithms and computing power, and significantly improve its transferability and interpretability. It can be said that big data is the enabling factor of the new generation of artificial intelligence and is crucial to promoting the development of artificial intelligence.

Lv Bin: At present, artificial intelligence technology is accelerating its penetration into the military field. Military intelligence has become the core driving force of a new round of military reforms, profoundly changing the winning mechanism, force structure and combat methods of future wars. Military intelligence is not just a simple superposition of artificial intelligence and military, but also a systematic description of the coordinated operation of people, equipment and combat methods under a new combat form. The key to the operation of the system is inseparable from the efficient acquisition, integration, analysis and interaction of data. In the future, with the full penetration of data into the military field, deep interaction between man and machine, and the deep combination of machine intelligence and human wisdom, autonomous perception, autonomous analysis, autonomous decision-making and autonomous strikes centered on data and centered on data analysis and processing will be realized.

Luo Wei: We should also look at the relationship between the two from the perspective of the historical stage of our military construction and development. At present, our military has made significant progress in the transformation of military reforms with Chinese characteristics, but the task of mechanization construction has not been completed, and the level of informatization needs to be improved urgently. The development of military intelligence must not be a “castle in the air”. It must be built on the basis of mechanization and informatization, and at the same time, it must also focus on using intelligence to drive mechanization and informatization construction. The integrated development of the “three transformations” will be a significant feature of our military construction and development at present and in the future. Through the comprehensive sharing and efficient interaction of various types of data, the information data flow of the integration of the “three transformations” will be opened up, which will help to build the underlying channel of the integration of the “three transformations” and effectively enhance the quality and efficiency of the development of military intelligence.

Major military powers are rushing to build military big data

Reporter: Accelerating the development of military intelligence has become a common practice for the world’s military powers to seize the commanding heights of future military competition. So what is the progress of these countries in promoting the construction of military big data and the development of military intelligence? Please give a brief introduction.

Liu Linshan: Winning by technology has always been the main means for Western countries to seek advantages in the world. With the advent of the era of big data and artificial intelligence, major Western countries regard it as a “battlefield” and have successively introduced a series of strategic measures to strengthen overall planning. As early as March 2012, the US government issued the “Big Data Research and Development Program Initiative”, which advocated the joint efforts of six government departments and agencies including the Department of Defense to jointly promote the development of big data collection, storage, management and analysis technologies. In 2016, the “National Artificial Intelligence Research and Development Strategic Plan” was formulated to clarify the investment direction and focus of the United States in the field of artificial intelligence. In 2018, the “Department of Defense Artificial Intelligence Strategy” was issued, which proposed artificial intelligence development goals and measures, and unified planning and deployment of military applications of artificial intelligence. This year, a new version of the “Department of Defense Cloud Strategy” was formulated to accelerate the development of big data and artificial intelligence technologies and promote data sharing through unified supervision of the US military’s cloud construction. In order to coordinate the development of artificial intelligence technology and combat applications of the US military, the US Department of Defense has also established a joint artificial intelligence center to integrate relevant resources and plan investments. With the launch of the US Department of Defense’s “Military Cloud” 2.0 system, the US military has acquired data processing capabilities such as high-speed battlefield video processing, voice recognition, complex electromagnetic environment perception, and decryption, reducing processing time to one thousandth of traditional methods. Russia has formulated the “Russian Federation Science and Technology Development Strategy”, which prioritizes big data, machine learning, and artificial intelligence in scientific and technological innovation, and the Russian military has formulated the “Military Robot Technology and Application Development Plan”. The United Kingdom has listed big data, robots, and autonomous systems as eight priority development technologies, and the military has set up a special artificial intelligence laboratory to focus on the research of artificial intelligence and defense data science. France’s digital roadmap clearly states that big data is a strategic high-tech that must be strongly supported in the future. The military has formulated the “Artificial Intelligence and Innovation Roadmap”, which lists intelligence, decision-making, human-machine collaboration, robots, and cyber warfare as key development areas.

Luo Wei: In addition to strategic layout, major Western countries have continued to increase their research and development efforts in big data and artificial intelligence technologies to accelerate the transformation of technological achievements into combat capabilities. Since 2012, the US Department of Defense and its subordinate departments have implemented a series of big data research and development projects represented by the “X Data” project and the “Insight” project, mainly involving big data analysis and mining, rule discovery, data-driven model calculation, data visualization and other fields. Among them, the “X Data” project aims to develop new computing technologies and open source software tools for big data processing and analysis; the “Insight” project aims to integrate massive data from various sensors to form a comprehensive battlefield situation, quickly identify the source and degree of threats, and enhance the decision-making ability of troop commanders and staff. While continuing to promote technological development, some of the US military’s achievements have begun to form combat capabilities, among which the most typical is the “Algorithm Warfare” project. The project was launched in April 2017 and aims to use artificial intelligence algorithms to find targets of interest from massive video data. The relevant achievements have been deployed to multiple US military departments since December of that year, and their video intelligence analysis and processing capabilities have been improved.

Big data will play a vital role in future wars

Reporter: At present, military high-tech with information technology as the core is changing with each passing day, which is driving the evolution of war into information warfare, and intelligent warfare is beginning to emerge. So what role will big data play in future information and intelligent warfare? Please briefly explain.

Liu Linshan: In the future, both the continuously developing information warfare and the rapidly advancing intelligent warfare will be driven by data, which is mainly reflected in the following three aspects:

First, the understanding of battlefield situation depends on the collection and processing of massive data. Whoever can penetrate the “fog of war” and more accurately and comprehensively understand the enemy’s situation and the combat environment will be able to seize the initiative in the war. On the battlefield of the future, data is the basis for restoring the battlefield situation and forming the enemy’s situation. It not only includes the results of our own reconnaissance, surveillance, and intelligence activities, but also includes massive geographic information data, human social culture data, and social media data. At present, it is not easy to comprehensively collect and process this data, because with the continuous acceleration of the development of networked informatization in the whole society and the increasing popularity of various digital devices, the amount of data in the whole society continues to grow exponentially. Statistics show that in 2013, the total amount of data in human society was about 4.4 trillion GB, and by 2020, this number will grow to 44 trillion GB. With such a large amount of data, without the support of advanced big data collection and processing technology, the value of the data cannot be reflected, and the overall picture of the battlefield cannot be recognized.

Secondly, the realization of combat functions depends on the analysis, distribution and utilization of data. Combat functions usually include intelligence, command and control, firepower strikes, battlefield mobility, combat support, etc. Among them, the intelligence function focuses on collecting, compiling and pushing data, command and control focuses on integrating, processing and distributing data, and firepower strikes, battlefield mobility and combat support generate new status data while utilizing data. It can be said that the performance of combat functions and the implementation of combat operations are the process of data recycling. The smoother and faster the data circulation of one party is, the more significant the combat effect of that party will be.

Third, the evolution of joint operations depends largely on the level of data sharing. The U.S. military is currently evolving from joint operations to multi-domain operations, which is manifested in the development of the coordination of combat capabilities between military services to the aggregation of combat capabilities in various combat domains, thereby achieving a closer and more precise use of forces across military services on the basis of lower-level forces. The premise for achieving this goal is to use the “cloud” as a means to achieve the sharing of all combat data.

Lü Bin: As the role of data in future wars becomes more and more obvious, the characteristics of data weaponization will also become increasingly apparent. “Data warfare” in which one side prevents the opponent from obtaining its own data, prevents the opponent from forming a comprehensive situation, or creates and spreads false data, misleads the judgment of the opponent’s intelligent system, and hinders the opponent’s combat function will become an important combat style.

Actively embrace the era of military big data

Reporter: When it comes to future intelligent warfare, its important feature is that the tempo of confrontation has significantly accelerated. The winning factor has changed from being able to win to being fast. Whoever seizes the initiative will have the initiative in the war. As we welcome the arrival of the era of military big data, how should we understand its key role in future intelligent warfare?

Liu Linshan: The key to winning by speed is how to shorten the “OODA loop” chain as soon as possible. From the perspective of “observation”, big data technology can effectively help commanders fully grasp the situation, capture subtle changes, and discover major signs. From the perspective of “judgment”, the use of big data technology to analyze the correlation of multi-source data can accurately judge the battlefield situation, break the “battlefield fog”, and significantly enhance the commander’s judgment and acumen. From the perspective of “decision-making”, the real-time fusion processing and visualization of data such as enemy situation, our situation and battlefield environment can help commanders accurately control the battlefield situation in real time, make decisions quickly, adjust deployments quickly, and gain the speed advantage of planning and decision-making. From the perspective of “action”, based on real-time online big data analysis, commanders can accurately plan tasks, calculate troops and firepower, and accurately issue combat orders, so as to dynamically and efficiently control troop actions and accurately evaluate combat effectiveness.

Lu Bin: In addition to bringing revolutionary impact to command and control efficiency, big data will also accelerate the transformation of combat command system and lead the transformation of command decision-making mode. The first is the flattening of command system. Under the big data environment, information systems are seamlessly connected, cloud computing provides powerful computing power, blockchain technology improves system security, and the command structure can be changed from “tree-like” to “net-like”, providing a material basis for streamlining command structure and realizing flat command. The second is the jointness of military systems. Based on the highly shared data pool of big data, various types of data are integrated to ensure the consistency of information acquisition. Unified network infrastructure, consistent data structure and data exchange standards ensure effective interconnection and interoperability between various forces and elements, creating conditions for the formation of an integrated joint combat system. The third is the simplification of troop organization. The collection and processing of massive intelligence information is transmitted to the cloud through the information system network for processing, which can free commanders from heavy information processing and realize the streamlined action organization with efficient operation under the support of a large system.

Luo Wei: Military big data also plays an important enabling role in improving the level of equipment intelligence, promoting the birth of intelligent weapons and equipment, and improving intelligent support capabilities. As the “oil” of the new era, big data will be the life source and aorta of intelligent equipment. The rapid acquisition, fine processing and precise distribution of data based on super computing power are multipliers for the combat effectiveness of intelligent equipment. With the gradual breakthrough of key technologies, data equipment closely combined with data resources, computing resources, big data management and analysis systems will appear on the battlefield in the future, including giant platforms used to support military combat command, equipment management, and logistics distribution, as well as small and medium-sized data equipment based on aircraft carriers, aircraft, ships, submarines, vehicles, etc., and also micro-data equipment embedded in unmanned combat, single-soldier combat, and precision guidance systems. These data equipment will play the role of “smart engine” in future intelligent combat, training, logistics support and other operations. At the same time, it should be emphasized that big data going to the battlefield may change equipment and the form of war, but it cannot change the decisive factors of war. In the era of intelligent warfare, the key factor that determines the outcome of war is still people. Big data and artificial intelligence technology cannot completely replace people, and cannot change the decisive position of people in war.

現代國語：

編者按為深入貫徹習主席所做的「推動實施國家大數據戰略」決策部署，加速軍事智能化發展，我軍上下已開始積極行動起來，軍事大數據的創新應用必將前所未有地激發數據活力、釋放資料價值、產生倍增效應，以開啟未來制勝之門。

為積極適應新情勢新要求，推動大數據與國防建設深度融合，促進軍事大數據研究領域專家交流合作，由軍事科學院主辦、主題為「軍事大數據推動軍事智慧化發展」的第二屆軍事大數據論壇，於今年8月22日至23日在京舉行。來自軍委機關、各戰區、各軍兵種、軍事科學院、國防大學、國防科技大學等單位的500餘名領導、專家和代表參加，圍繞軍事大數據發展的前沿、共性、熱點問題進行了探討交流。

論壇期間，本報記者就軍事大數據的相關議題，訪問了軍事科學資訊研究中心主任劉林山研究員、副主任呂彬研究員、某研究室主任羅威副研究員。

軍事大數據具有強烈對抗性

記者：我國目前正大力推動實施國家大數據戰略。大數據正廣泛應用於經濟、政治、安全和社會管理等許多領域，體現出前所未有的巨大價值。那麼，與民用大數據相比，軍事大數據有著怎樣的內涵特徵？

劉林山：隨著大數據技術與應用的發展，軍事大數據已突破過去軍事數據的概念範疇，成為以海量軍事數據資源為基礎、以數據智能處理分析技術為核心、以軍事領域廣泛應用需求為牽引的一系列活動的統稱。鑑於軍事活動的特殊性，軍事大數據除具有民用大數據典型的資料規模大、內容種類多、處理速度高、價值密度低等特徵外，還具有「一超一高一強」的特性。

「一超」即超複雜性。指資料來源於陸、海、空、天、電、網等多個空間，資訊維度較高，非結構化特徵較明顯，資料關係較複雜。 「一高」即高安全性，指面臨的威脅複雜，包括敵方的偵察竊取，己方洩密失密、系統漏洞，遭敵「軟」「硬」手段打擊等，可用性削弱或喪失風險更大。 「一強」即強對抗性，指資訊獲取與反獲取手段的博弈對抗、資料迷霧偽裝欺騙現象普遍存在，真假資料錯綜複雜，對資料真偽辨別能力要求極高。

呂彬：在這裡需要突顯軍事大數據「強對抗性」的特徵。由於軍事大數據是對抗環境下的數據，數據品質差，價值密度低，通常具有不確定性、不完全性和虛假欺騙性。我們知道，現階段的人工智慧主要是建立在數據驅動的機器學習之上的。而機器學習需要樣本數據，但目前戰爭是小樣本數據，未來戰爭甚至沒有樣本數據。此外，軍事行動的特點是“人在迴路”，人的活動很難用大數據經典方法學習預測。這就使得軍事大數據在應對小樣本資料學習、不完全不確定資訊下的博弈、複雜環境下的場景建模與理解等問題方面，比民用大數據要困難得多，遇到的挑戰要大多，必須採用新的理論、新的方法、新的技術去解決。

軍事大數據與軍事智慧化相輔相成

記者：黨的十九大報告強調，要「加速軍事智慧化發展，提升基於網路資訊體系的聯合作戰能力、全域作戰能力」。對於軍事大數據與軍事智能化的關係，該怎麼理解？

劉林山：回顧人工智慧的發展歷程，自1956年達特茅斯會議首次提出以來，經歷了推理期、知識期、學習期三次高潮和兩次低潮。科學家曾試圖透過邏輯推理、專家系統等方式來「製造」人類智能，但日本智能計算機的研發失敗、美國史丹佛大學人類常識知識百科全書的沒落等，使得這些路徑難以為繼。而大數據的出現，為人工智慧的發展提供了一條新道路，人們開始從如何「製造」智能轉向如何「學習」智能。與傳統的希望透過規則、邏輯和知識來實現推理學習不同，透過機器學習，從大數據中去洞悉海量資料隱藏的規律，可有效實現資料驅動下的人工智慧。 2016年3月，Google公司開發的「阿爾法狗」圍棋機器人之所以能擊敗世界冠軍李世石，正是基於對3000萬盤棋譜數據的神經網路深度學習。除此之外，大數據技術的發展還可以彌補人工智慧在演算法、算力方面的不足，顯著提高其可遷移性和可解釋性。可以說，大數據是新一代人工智慧的賦能要素，對於推動人工智慧發展至關重要。

呂彬：目前，人工智慧技術正加速滲透到軍事領域。軍事智慧化已成為新一輪軍事變革的核心驅動力，深刻改變未來戰爭的致勝機理、力量結構和作戰方式。軍事智慧化不僅是人工智慧與軍事的簡單疊加，也是人、裝備和作戰方式在新的作戰形態下協同運作的體系化描述。體系運作的關鍵離不開資料的高效獲取、融合、研判、互動。未來，隨著數據向軍事領域全方位滲透，人機深度交互，機器智能與人類智慧深度結合，將實現以數據為中心、以分析處理數據為中樞的自主感知、自主分析、自主決策、自主打擊。

羅威：我們也要站在我軍建設發展所處歷史階段來看兩者之間的關係。目前我軍中國特色軍事變革取得重大進展，但機械化建設任務尚未完成，資訊化程度亟待提升。軍事智慧化發展絕不能是“空中樓閣”，要建立在機械化和資訊化的基礎上，同時也要注重用智慧化來牽引機械化和資訊化建設。這「三化」融合發展，將是當前及今後一段時期我軍建設發展的顯著特徵。透過各類數據的全面共享和高效交互，打通「三化」融合的資訊資料流，有助於建構「三化」融合的底層通道，實際增強軍事智能化發展的品質效益。

主要軍事強國紛紛搶灘軍事大數據建設

記者：加速推動軍事智慧化發展，目前已成為世界軍事強國搶佔未來軍事競爭制高點的共同做法。那麼這些國家在推動軍事大數據建設與軍事智慧化發展的進展如何？請簡單介紹一下。

劉林山：科技制勝一直是西方國家在世界上尋求優勢的主要手段。隨著大數據和人工智慧時代的到來，西方各主要國家將其視為“兵家必爭之地”，先後出台一系列戰略舉措，強化統籌佈局。美國政府早在2012年3月就發布了《大數據研究與發展計畫倡議》，倡議聯合國防部在內的6個政府部門和機構，共同推動大數據收集、儲存、管理、分析技術的發展。 2016年制定《國家人工智慧研發戰略規劃》，明確美國在人工智慧領域的投資方向與重點。 2018年發布《國防部人工智慧戰略》，提出人工智慧發展目標和舉措，對人工智慧的軍事應用進行了統一規劃和部署。今年制定新版《國防部雲端戰略》，透過對美軍雲建設的統一監管，加速大數據與人工智慧技術發展，促進資料共享。為統籌美軍人工智慧技術發展與作戰運用，美國防部也特別成立了聯合人工智慧中心，整合相關資源與計畫投資。隨著美國防部「軍事雲」2.0系統上線運行，美軍已具備高速戰場視訊處理、語音辨識、複雜電磁環境感知、解密等資料處理能力，使處理時間縮短到傳統方法的數千分之一。俄羅斯制定了《俄羅斯聯邦科技發展戰略》，將大數據、機器學習和人工智慧作為科技創新優先方向，俄羅斯軍隊則制定了《軍用機器人技術和應用發展規劃》。英國將大數據、機器人和自主系統列入八項優先發展技術，軍方專門組成了人工智慧實驗室，聚焦人工智慧和國防數據科學的研究。法國的數位化路線圖明確大數據是未來必須大力支持的戰略性高新技術，軍方制定有《人工智慧與創新路線圖》，將情報、決策、人機協同、機器人與網路戰列為重點發展領域。

羅威：除了戰略佈局之外，西方主要國家還持續加強數據和人工智慧技術的研發力度，加速技術成果轉化為作戰能力。自2012年以來，美國防部及其所屬各部門就實施了以「X數據」項目、「洞察」項目為代表的一系列大數據研發項目，主要涉及大數據分析挖掘、規則發現、數據驅動模型計算、數據視覺化等領域。這其中，「X數據」專案旨在為大數據處理和分析開發新型運算技術和開源軟體工具；「洞察」專案旨在融合來自各類感測器的大量數據，形成全面戰場態勢，快速識別威脅來源和威脅程度，增強部隊指揮和參謀人員的決策能力。在持續推進技術發展的同時，美軍部分成果開始形成作戰能力，其中最為典型的就是「演算法戰」計畫。該計畫於2017年4月啟動，旨在採用人工智慧演算法從大量的視訊資料中發現感興趣的目標，相關成果從當年12月開始陸續部署美軍多個部門，其視訊情報分析處理能力得到提升。

大數據在未來戰爭中將扮演重要角色

記者：目前，以資訊科技為核心的軍事高新科技日新月異，正推動戰爭形態朝向資訊化戰爭演變，智慧化戰爭初露端倪。那麼在未來資訊化、智慧化戰爭中，大數據將扮演什麼樣的角色？請簡要說明一下。

劉林山：未來，無論是持續發展的資訊化戰爭形態，或是迅速推進的智慧化戰爭形態，都是由數據驅動的，主要表現在三個方面：

首先，戰場態勢的理解依賴於海量資料的收集、處理。誰能穿透“戰爭迷霧”，更準確、全面地洞悉敵我態勢和作戰環境，誰就能在戰爭中搶佔先機。在未來戰場上，資料就是還原戰場情況、形成敵我態勢的基礎，它不但包括己方偵察、監視、情報活動的成果，也包括海量的地理資訊資料、人類社會文化資料、社群媒體資料。目前，要全面收集、處理這些數據並不容易，因為隨著全社會網路化資訊化發展的持續加速，各類數位設備的日益普及，全社會數據量持續呈現指數級增長。有統計數據表明，在2013年，人類社會的數據總量約為4.4兆GB，到2020年，這一數字將增長到44兆GB。如此之大的數據量，如果沒有先進的大數據收集、處理技術支持，數據的價值就無從體現，戰場的全貌也就無從認知。

其次，作戰功能的實現取決於資料的分析、分發和利用。作戰功能通常包括情報、指揮控制、火力打擊、戰場機動、作戰保障等。這其中，情報功能著重於收集、整編、推播數據，指揮控制著重於融合、處理、分發數據，而火力打擊、戰場機動、作戰保障在利用數據的同時，也產生新的狀態數據。可以說，作戰功能的發揮、作戰行動的實施就是資料循環利用的過程。哪一方的資料循環得越順暢、越快速，哪一方的作戰效果就越顯著。

其三，聯合作戰樣式的演進在很大程度上取決於資料共享程度。當前美軍正由聯合作戰向多域作戰方向演進，表現為從軍兵種間作戰能力協同向各作戰域作戰能力聚合的方向發展，從而在更低層級部隊基礎上實現跨越軍兵種的更緊密、更精確的力量運用。達成此目標的前提，就是以「雲」為手段，實現所有作戰資料的共享。

呂彬：由於數據在未來戰爭中的作用愈發明顯，數據武器化的特性也將日益顯露。一方防止對手獲取己方數據、阻止對手形成全面態勢，或製造散佈虛假數據、誤導對手智能係統的判斷、阻礙對手作戰功能實現的“數據戰”，將成為重要的作戰樣式。

主動擁抱軍事大數據時代

記者：提到未來智慧化戰爭，其重要特徵就是對抗節奏明顯加快，制勝機理由以能製勝轉變為以快制勝，誰掌握了先機，誰就掌握了戰爭的主動權。我們迎接軍事大數據時代的到來，該如何認識其在未來智慧化戰爭中的關鍵角色？

劉林山：以快速致勝的關鍵，在於如何盡快縮短「OODA環」循環鏈。從「觀察」環節來看，大數據技術能有效幫助指揮員全面掌握狀況、捕捉細微變化、發現重大徵候。從“判斷”環節看，運用大數據技術對多來源數據關聯分析，可以準確研判戰場形勢、破除“戰場迷霧”，顯著增強指揮員的判斷力和敏銳性。從「決策」環節來看，敵情、我情和戰場環境等數據的即時融合處理和視覺化展示，能夠幫助指揮官即時準確掌控戰場態勢，快速作出決策、快速調整部署，贏得籌劃決策的速度優勢。從「行動」環節來看，基於即時線上大數據分析，指揮員能夠精準規劃任務、精細運算兵力火力、精確下達作戰指令，從而動態高效調控部隊行動、準確評估作戰效果。

呂彬：大數據除了為指揮控制效能帶來革命性影響外，還將加速作戰指揮體系變革、引領指揮決策模式轉變。首先是指揮體系扁平化。大數據環境下，資訊系統無縫鏈接，雲端運算提供強大運算能力，區塊鏈技術提升系統安全性，指揮結構可由“樹狀”變為“網狀”，為精簡指揮結構、實現扁平化指揮提供了物質基礎。其次是軍兵種體系聯合化。基於大數據高度共享的資料池，將各類資料融合在一起，確保取得資訊的一致性。統一的網路基礎設施、一致的資料結構與資料交換標準，確保各力量、各要素之間有效互聯互通互通，為形成一體化聯合作戰體系創造條件。第三是部隊編組精簡化。海量情報資訊的收集、處理透過資訊系統網路傳輸至雲端進行處理，能夠使指揮人員從繁重的資訊處理中解放出來，實現大體系支撐下高效運作的精簡化行動編組。

羅威：軍事大數據在提高裝備智慧化水準、催生智慧化武器裝備和提高智慧化保障能力等方面也具有重要的賦能作用。作為新時代的“石油”，大數據將是智慧化裝備的生命源泉和大動脈，基於超強算力的數據快速獲取、精細處理和精準分發是智能化裝備作戰效能發揮的倍增器。隨著關鍵技術的逐步突破，未來戰場將出現由資料資源、運算資源、大數據管理與分析系統等緊密結合的資料裝備，既包括用於支撐軍兵種作戰指揮、裝備管理、後勤配給的巨型平台，也包括以航空母艦、飛機、艦艇、潛艦、車輛等為載體的中小型資料裝備，還包括嵌入無人作戰、單兵作戰、精確導引系統的微小型資料裝備。這些資料裝備，將在未來智慧化作戰、訓練、後勤保障等行動中發揮「智慧引擎」作用。同時要強調的是，大數據走向戰場可能改變裝備，可能改變戰爭形態，但不可能改變戰爭的決定因素。智慧化戰爭時代，決定戰爭勝負的關鍵因素仍是人。大數據和人工智慧技術不可能完全取代人，不能改變人在戰爭中的決定性地位。

中國原創軍事資源：http://www.81.cn/2019zt/2019-09/06/content_9623888.htm