Scholarlink

<p>Managing critical health crises such as the COVID-19 pandemic necessitates minimizing direct contact between healthcare workers and infected individuals to reduce the risk of viral transmission. One effective solution is the deployment of autonomous robots capable of performing logistical tasks, such as delivering medication and food, as well as monitoring environments without human intervention. The robot operating system (ROS) provides a flexible platform for developing such robots, offering modular algorithm packages tailored for various functions. Indoor autonomous navigation, in particular, requires accurate sensor configurations for precise localization within confined environments. This study proposes a sensor fusion approach combining Marvelmind, light detection and ranging, inertial measurement unit, and rotary encoder data, integrated using the extended Kalman filter (EKF) to enhance localization accuracy. EKF, a robust algorithm for estimating the state of nonlinear systems from noisy measurements, is implemented via ROS to support autonomous navigation. Experimental results show that the robot, equipped with the proposed sensor configuration and EKF, successfully navigated a square track with a mean square error of 0.110108. Furthermore, it autonomously reached three target destinations with an average positional deviation of 0.2471 meters. In comparative tests, the system without Marvelmind recorded an average deviation of 0.441431 meters when navigating to three rooms, while integration of Marvelmind reduced the deviation significantly to 0.172779 meters. These results highlight the effectiveness of the proposed system in achieving reliable and accurate indoor autonomous navigation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhanced localization for autonomous smart medical robots: Marvelmind-lidar-IMU fusion with extended Kalman filter

  • 期刊论文

  • 发表日期：

    2026年3月10日

查看全文
  • Arjon Turnip,
  • Darmawan Hidayat,
  • Dwi Esti Kusumandari,
  • Endra Joelianto,
  • Ubaidillah

摘要

Managing critical health crises such as the COVID-19 pandemic necessitates minimizing direct contact between healthcare workers and infected individuals to reduce the risk of viral transmission. One effective solution is the deployment of autonomous robots capable of performing logistical tasks, such as delivering medication and food, as well as monitoring environments without human intervention. The robot operating system (ROS) provides a flexible platform for developing such robots, offering modular algorithm packages tailored for various functions. Indoor autonomous navigation, in particular, requires accurate sensor configurations for precise localization within confined environments. This study proposes a sensor fusion approach combining Marvelmind, light detection and ranging, inertial measurement unit, and rotary encoder data, integrated using the extended Kalman filter (EKF) to enhance localization accuracy. EKF, a robust algorithm for estimating the state of nonlinear systems from noisy measurements, is implemented via ROS to support autonomous navigation. Experimental results show that the robot, equipped with the proposed sensor configuration and EKF, successfully navigated a square track with a mean square error of 0.110108. Furthermore, it autonomously reached three target destinations with an average positional deviation of 0.2471 meters. In comparative tests, the system without Marvelmind recorded an average deviation of 0.441431 meters when navigating to three rooms, while integration of Marvelmind reduced the deviation significantly to 0.172779 meters. These results highlight the effectiveness of the proposed system in achieving reliable and accurate indoor autonomous navigation.