机器人即服务(RaaS)的发展与安全规制  

The Development and Safety Regulation of Robot as a Service RaaS

作　　者：金文琴 左晓栋 Jin Wenqin;Zuo Xiaodong(School of Cyber Science and Technology,University of Science and Technology of China,Hefei Anhui,230026;School of Public Affairs,University of Science and Technology of China,Hefei Anhui,230026)

机构地区：[1]中国科学技术大学网络空间安全学院,安徽合肥230026 [2]中国科学技术大学公共事务学院,安徽合肥230026

出　　处：《工业信息安全》2024年第2期62-70,共9页Industry Information Security

基　　金：2023年国家社科基金重大项目“人工智能时代大国跨境数据战略竞争与中国应对研究”(批准号:23&ZD335)。

摘　　要：机器人即服务(Robot as a Service,RaaS)结合了云计算、人工智能、共享服务和机器人技术,能在高效满足服务需求的同时降低使用和部署机器人设备的成本,因而这种独特的服务模式自提出就得到了广泛关注。随着RaaS市场规模的不断扩大,其自身技术缺陷以及人为不当使用导致了一系列的安全风险。此外ChatGPT等生成式人工智能技术的兴起解锁了一种新的机器人应用范式,自主、互连和智能化将成为机器人发展的必然趋势,这也会给RaaS带来新的安全风险。本文首先介绍了RaaS的兴起和发展,明确了其基本概念及主要特征,并从体系结构的角度分析了这种服务模式,然后结合人工智能阐述了RaaS的安全风险和对其进行规制的必要性,最后从多个层面提出了风险控制措施建议。Robot as a Service(RaaS)combines cloud computing,artificial intelligence,shared services,and robotics to efficiently meet service needs while reducing the cost of using and deploying robotic devices,and thus this unique service model has gained widespread attention since its introduction.As the size of the RaaS market continues to expand,its own technological shortcomings and improper human use have led to a series of security risks.In addition,the rise of generative AI technologies such as ChatGPT has unlocked a new paradigm of robotics applications,where autonomy,interconnectivity and intelligence will become an inevitable trend in the development of robots,which will also bring new security risks to RaaS.This paper firstly introduces the rise and development of RaaS,clarifies its basic concepts and main features,and analyses this service model from the perspective of architecture,then describes the security risks of RaaS and the necessity of regulating it in the context of AI,and finally puts forward proposals for risk control measures from multiple levels.

关 键 词：机器人即服务 生成式人工智能 安全规制 

分 类 号：TP2[自动化与计算机技术—检测技术与自动化装置]