There are many issues to be studied in the field of robotics , involving multiple disciplines such as computer science, sensors , human-computer interaction, and biosafety. Among them, environmental perception, autonomous localization, and motion control are the three key issues in robotics technology. The following will discuss these three points in detail.
Environmental perception
Currently, in indoor robot environments, autonomous environmental perception technology for mobile robots, which mainly uses LiDAR and is supplemented by other sensors, is relatively mature. However, in outdoor applications, due to the variability of the environment and changes in lighting, the task of environmental perception is much more complex and has higher requirements for real-time performance. This makes multi-sensor fusion a major technical task for robot environmental perception.
Using a single sensor for environmental perception often has its own insurmountable weaknesses. However, by effectively fusing multiple sensors and redundancy and complementarity of information from different sensors, robots can cover almost all spatial detection and comprehensively improve their perception capabilities. Therefore, using lidar sensors in combination with ultrasonic, depth camera, and anti-fall sensors to obtain distance information to enable robots to perceive their surroundings has become a hot research topic for scholars around the world.
Using multiple sensors for environmental perception technology brings challenges such as the synchronization, matching, and communication of multi-source information, necessitating research into methods and technologies for cross-modal and cross-scale information registration and fusion from multiple sensors. However, in practical applications, using more types of sensors is not always better. For specific robot applications in different environments, the validity of data from each sensor and the real-time performance of computation must be considered.
Autonomous positioning
For mobile robots to achieve autonomous walking, positioning is one of the core technologies they need to master. Currently, GPS can provide high accuracy in global positioning, but GPS has certain limitations. In indoor environments, GPS signals may be weak, which can easily lead to loss of location.
In recent years, SLAM technology has developed rapidly, improving the localization and mapping capabilities of mobile robots. SLAM is an abbreviation for Simultaneous Localization and Mapping, first proposed by Hugh Durrant-Whyte and John J. Leonard in 1988. SLAM is more accurately described as a concept than an algorithm; it is defined as a collective term for methods that solve the problem of "a robot starting from an unknown location in an unknown environment, determining its own position and orientation by repeatedly observing map features (such as corners, pillars, etc.) during movement, and then incrementally building a map based on its own position, thereby achieving the goal of simultaneous localization and mapping."
Path planning
Path planning is also an important branch of robotics research. Optimal path planning is based on one or more optimization criteria (such as minimizing work cost, shortest travel route, and shortest travel time) to find an optimal path from the initial state to the target state in the robot's workspace that avoids obstacles.
Depending on the degree of understanding of environmental information, robot path planning can be divided into global path planning and local path planning.
Global path planning involves planning a path for a robot within a known environment. The accuracy of the path planning depends on the accuracy of the environment data. Global path planning can find the optimal solution, but it requires accurate prior knowledge of the environment. When the environment changes, such as the appearance of unknown obstacles, this method becomes ineffective. It is a form of pre-planning, therefore it does not require high real-time computing power from the robot system. Although the planning result is global and relatively optimal, it has poor robustness to errors and noise in the environment model.
Local path planning, on the other hand, deals with environments where the information is either completely unknown or partially known. It focuses on considering the robot's current local environment information to give the robot good obstacle avoidance capabilities. Sensors are used to detect the robot's working environment to obtain information such as the position and geometry of obstacles. This planning requires the collection of environmental data and the ability to correct the dynamic updates of the environmental model at any time. Local planning methods integrate environmental modeling and search, requiring the robot system to have high-speed information processing and computing capabilities, high robustness to environmental errors and noise, and the ability to provide real-time feedback and correction of planning results. However, due to the lack of global environmental information, the planning results may not be optimal, and it may even be impossible to find the correct or complete path.
There is no fundamental difference between global path planning and local path planning. Many methods suitable for global path planning can also be used for local path planning after modification, and vice versa. Working together, the robot can better plan its path from the starting point to the destination.
What is the current status of sensing, localization, and path planning technologies?
To address the challenge of autonomous robot movement, numerous domestic companies are researching technologies such as environmental perception, autonomous localization, and path planning. Slamtec, a leading domestic company in robot positioning and navigation technology, already possesses relatively mature products for achieving autonomous robot movement, such as Apollo, which helps companies reduce R&D costs. The Apollo robot chassis is equipped with laser rangefinders, ultrasonic sensors, and anti-fall sensors, and a depth camera sensor is mounted on the chassis. Combined with its independently developed SLAMWARE autonomous navigation and localization system, the robot achieves autonomous mapping, localization, and navigation capabilities.
When Apollo is in an unknown environment, it can construct a high-precision, centimeter-level map using SharpEdge™ fine-grained mapping technology without modifying the environment. This map boasts ultra-high resolution and eliminates the risk of error accumulation. Simultaneously, it utilizes the D* dynamic real-time path planning algorithm to find a path and move to the designated location, requiring no secondary optimization or modification, directly meeting user expectations.
In addition, based on a purely software-based approach, no additional auxiliary installations are required. Apollo can be pre-programmed with routes, or virtual walls and virtual tracks can be set to prevent Apollo from entering a restricted work area.
When Apollo's battery is low during operation, it can support scheduled charging with autonomous navigation and positioning, and automatically return to the charging dock for charging.
In addition, Apollo's expansion interfaces integrate network ports, power supply interfaces, and various control interfaces to facilitate rapid development and expansion by users. Apollo can communicate with external devices via wired network or Wi-Fi, and its built-in battery can power itself and external expansion modules. Users can control the entire Apollo and its upper-layer expansion modules through various control interfaces.
In conclusion, in recent years, governments around the world have attached great importance to the development of robotics technology and invested a lot of resources to encourage robotics companies to innovate and forge ahead. It is believed that in the future, robots will become an important part of people's daily lives and lead people to a more convenient era!
Disclaimer: This article is a reprint. If it involves copyright issues, please contact us promptly for deletion (QQ: 2737591964). We apologize for any inconvenience.
