Development of a low-cost modular snake-like robot with 2-DOF modules for rescue operations in collapsed environments with fast communication
Locating trapped or lost individuals is critical in search and rescue, directly impacting survival rates and responder safety. Rapid victim detection enhances response efficiency and enables timely rescue actions. Integrating advanced detection with mobile robots minimizes human risk, with snake rob...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SAGE Publishing
2025-07-01
|
| Series: | International Journal of Advanced Robotic Systems |
| Online Access: | https://doi.org/10.1177/17298806251360659 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Locating trapped or lost individuals is critical in search and rescue, directly impacting survival rates and responder safety. Rapid victim detection enhances response efficiency and enables timely rescue actions. Integrating advanced detection with mobile robots minimizes human risk, with snake robots offering a promising solution due to their flexible, modular design for navigating complex and confined environments. This article focuses on developing a snake-like robot (His Hands version-1/HH-I) for search and rescue missions. This article concentrates on two aspects of the robot: one is the design and development of the mechanism for traversal in a cluttered environment, and the other is establishing a fast communication system for sending victim detection information (based on a previously developed victim detection model) to rescue victims. The highlight of the work is the robust design made using CAD. The model was then developed and tested in real time, and most of the components in the prototype were built based on 3D printing and other in-house fabrications. The tracks on all four sides ensure self-recovery after rollovers, with a maximum slope climbing capacity of 40°. Experimental tests demonstrated a top speed of 6.8 cm/s on flat terrain and successful traversal on rubble, sand, and gravel surfaces. Victim detection is achieved using a transfer learning-based ResNet-50 model with 97.2% accuracy in an average detection time of 11.5 s. Then, the victim detection information is sent to the rescue teams’ mobile app with IoT using the IBM cloud and node-red. |
|---|---|
| ISSN: | 1729-8814 |