Intelligent Security Robot Overview
Security industry, with the security needs of modern society emerged in the industry. It can be said, as long as there is crime and there is social unrest, the security industry will exist and develop. It turns out that social crime is often not due to the development of social, economic prosperity and reduction.
Throughout the course of the development of the security industry, mainly around the video monitor constantly upgrading reform, it has experienced a total of five times change. "Visible, far ahead, see clearly" has been a major factor in promoting progress. Since the product of AI technology introduced in 2015, making the industry take the initiative to identify the transition from passive monitoring, "could understand" has become the main theme of video surveillance products. 2016, 2017 through the joint action of national efforts to build a positive and innovative enterprise, in 2018 the smart security industry into the rapid landing stage, the technology began to really energize the industry. With the autopilot technology-related industrial chain development, security industry robots have begun to emerge in the market.
As shown in FIG overall security chain:
The basic status of security robot
Since 2013, China has become the world's largest robot market in the next 10--15 years have vast likely to remain in this position. In the R & D investment, China also made a lot of efforts. According to IDC forecasts: 2018 expenditures global robotics and unmanned solutions will reach $ 103.1 billion, of which robot technology spending will reach $ 94 billion, the total 2017-- more than 90% of the 2021 forecast expenditure. Among them, the industrial robot solutions will account for the largest share of expenditure robot more than 70%, followed by consumer service robots, and robots. Meanwhile, China's robot industry has a complete industrial chain. Therefore, the level of security guarding hundreds of millions of the market cake, in the environment of the country Skynet engineering and sharp engineering, combined with complete industrial chain of China robot, security robot can be said to stand on two industrial and trade outlets.
Security industry has a huge demand gap, it seems to be the new direction of service robots. Alternatively robot security video surveillance and security personnel to some extent, become an option no security field.
In some special work areas, with features intelligent robots can move, anomalies can effectively monitor the environment occur.
举个示例：电力线巡检机器人，可以搭载红外温度传感摄像头，对整个作业区域的电力设备进行监测，一旦出现异常情况，上报告警至监控平台。对电力设备的温度检测采用人工或者全线铺设广泛监测装置的方式，不是很好解决办法。人工巡检存在管理上风险，漏检、未检、巡检人身安全都是作业过程中存在的问题。全线铺设方式，一次投入的成本过大，变更场地不够灵活。采用智能巡检机器人，可以柔性投入，机器人按照设定的巡检路线进行作业，不存在漏检、未检的风险。同时机器人能够全天候作业，按照巡检任务全勤进行线路监控。
安防机器人类别
安防智能机器人根据使用场景可分为，两种主要的场景：封闭环境、开放环境。
封闭环境的应用领域主要有：机房巡检（数据中心、电信机房等）、电商仓库、电力巡检（变电站、输配线路、核电站等）。
开放环境的应用领域主要有：轨道交通、景区、工业园区、住宅小区等。
应用在开放环境中巡检机器人对导航的精准度和专业检测、报警能力有更高要求。使用的机器人智能检测基础能力包括：烟火检测、高温告警、异常声音告警等；巡逻机器人因使用场景丰富，所以需求空间更大，由于需要暴露在开放的室外，所以对导航稳定性、流畅性，本体鲁棒性等方面有很大的要求。巡检机器人完成的任务主要包括、人脸识别、人体检测、辆检测及识别、烟火检测、异常行为分析以及人证核验、语音交互与语义分析等。
技术及难点
智能安防机器人是实现主动安防的一次新突破，借助智能机器人的技术特征可达到巡检、巡逻、监控三者三位一体相辅相成。
智能安防机器人的核心技术包括本体的低速无人驾驶技术（由底盘技术、传感器组合和自主导航 SLAM 技术组成）、以计算机视觉为主的VSLAM技术、人工智能视频分析技术和网络传输、云平台管控相关技术。尤其是5G应用逐渐开始商用，将来智能巡检机器人借助5G特征可以完成更多任务。在落地应用中安防智能机器人有两大技术难点，一是如何保证 SLAM 技术在动态导航中的稳定性和准确性；二是如何保证在光照强弱不一、恶劣天气情况，以及多移动目标等复杂场景下，机器智能技术的可用性。
目前智能安防机器人还无法实现自学、自理、自决 等完全智能的能力， 但已经具备了目标检测、物体识别、环境感知、多模态交互等基本的分析判定功能，未来相关技术的突破还有赖于学术成果落地，以及市场的实际需求的推动。
安防智能机器人系统构成包括机器人、网络系统、机器人云平台、 可拓展设备， 以及充电坞 ，其中机器人本体近年来不断增加边缘计算能力，多采用 CPU+GPU 的芯片模式；网络传输方面多为视频传输需求，使用包括 WiFi 、移动互联网（4G、5G）、BLE通信、LoRA等多渠道传输。云平台主要作用是多机器人管理调度、应急控制、业务处理、视频分析、数据报表、安全态势分析、第三方数据平台接入等，是开放性合作的重要载体。安防智能机器人在落地部署方面 主要包括 机器人运行环境勘察 、 制定网络部署方案 、 制定机器人巡检 巡逻方案 、 机器人建图 ，以及机器人巡检巡逻路径 和 工作目标设定 。
Currently the actual delivery of the high demand for technical support team, need technical support staff for the various functions of the robot, various types of network characteristics, the technical team of the project implementation can be familiar with. Characteristics for different industries, different deployment need to know, the robot requires companies have a deeper insight into the industry. With the improvement of the overall system of the future stabilization of the robot, the deployment process will be more simple.