| Item type |
Symposium(1) |
| 公開日 |
2021-03-15 |
| タイトル |
|
|
タイトル |
Resource Utilization Prediction Model for SLAM Offload to Edge |
| タイトル |
|
|
言語 |
en |
|
タイトル |
Resource Utilization Prediction Model for SLAM Offload to Edge |
| 言語 |
|
|
言語 |
eng |
| 資源タイプ |
|
|
資源タイプ識別子 |
http://purl.org/coar/resource_type/c_5794 |
|
資源タイプ |
conference paper |
| 著者所属 |
|
|
|
Department of Information Science and Engineering, Faculty of Engineering, Shibaura Institute of Technology |
| 著者所属 |
|
|
|
Ales Inc |
| 著者所属 |
|
|
|
Department of Information Science and Engineering, Faculty of Engineering, Shibaura Institute of Technology |
| 著者所属(英) |
|
|
|
en |
|
|
Department of Information Science and Engineering, Faculty of Engineering, Shibaura Institute of Technology |
| 著者所属(英) |
|
|
|
en |
|
|
Ales Inc |
| 著者所属(英) |
|
|
|
en |
|
|
Department of Information Science and Engineering, Faculty of Engineering, Shibaura Institute of Technology |
| 著者名 |
Koki, Nagahama
Yoichi, Ishiwata
Midori, Sugaya
|
| 著者名(英) |
Koki, Nagahama
Yoichi, Ishiwata
Midori, Sugaya
|
| 論文抄録 |
|
|
内容記述タイプ |
Other |
|
内容記述 |
In recent years, the use of autonomous mobile robots has been increasing. SLAM technology is widely used for autonomous mobile robots to estimate their position and to create an environmental map at the same time. However, SLAM has a disadvantage because a load of various image processing and computation tasks is high. Also, in systems with extremely limited hardware resources, such as embedded systems, the execution speed is likely to be slowed down. Although various offloading methods have been proposed, they have not been applied to ROS+SLAM applications, and there are few examples of verification with robots. Therefore, in this study, we build a prediction model based on system resource information for a certain period associated with SLAM processing. Based on this resource usage prediction, we also propose a load-reducing method to offload data that is less frequently accessed on memory to the edges as needed. We construct a prediction model using the data obtain ed from the robot and describe the results of the evaluation. |
| 論文抄録(英) |
|
|
内容記述タイプ |
Other |
|
内容記述 |
In recent years, the use of autonomous mobile robots has been increasing. SLAM technology is widely used for autonomous mobile robots to estimate their position and to create an environmental map at the same time. However, SLAM has a disadvantage because a load of various image processing and computation tasks is high. Also, in systems with extremely limited hardware resources, such as embedded systems, the execution speed is likely to be slowed down. Although various offloading methods have been proposed, they have not been applied to ROS+SLAM applications, and there are few examples of verification with robots. Therefore, in this study, we build a prediction model based on system resource information for a certain period associated with SLAM processing. Based on this resource usage prediction, we also propose a load-reducing method to offload data that is less frequently accessed on memory to the edges as needed. We construct a prediction model using the data obtain ed from the robot and describe the results of the evaluation. |
| 書誌情報 |
Proceedings of Asia Pacific Conference on Robot IoT System Development and Platform
巻 2020,
p. 79-80,
発行日 2021-03-15
|
| 出版者 |
|
|
言語 |
ja |
|
出版者 |
情報処理学会 |
IPSJ-APRIS2020020.pdf (1.3 MB)
