Exploring Uncharted Terrain: Vision-Enabled Collaborative Robots

Exploring Uncharted Terrain: Vision-Enabled Collaborative Robots

1. Why were they created?

   • Is a great option for a first exploration after:
     • Natural Events
       • Earthquakes
       • Hurricanes
       • Attacks

2. Main benefits empirically evidenced by its use

   • Assists in primary exploration and rescuers in search and rescue missions
     • In the SAR context, collaborative robotics is important since it addresses larger areas, increasing the efficiency and precision in early detection.
     • Among the main developments in this area
       • Data collection
         • Inspection of areas devastated by a natural disaster
           • Victims identification
       • The most significant challenges in Search and Rescue (SAR) robotics.
         • Disorganized terrain
           • Limiting robotic exploration due to the presence of uneven terrain.
           • Technological challenges arise
             • Characterization of the environment
               • Autonomy in exploration
                 • Management of the communication

3. Main idea of the article

   • Develop a team of two collaborative robots (R1 and R2) and multisensory processing to perform autonomous climbing in irregular areas.
     • R2
       • Components
         • RGB-D sensor
           • For data acquisition.
     • R1 (Dr Robot Jaguar)
       • Components
         • 2DoF ramp
           • Capable of generating a bridge between the ground and the elevated area, allowing R2 to climb.
           • Visual headlights
             • ToF sensors
               • To assist the R2 in ascending and descending autonomous.
     • Algorithms
       • Based on vision and multisensory processing have been implemented to carry out autonomous collaborative ascent, which consists of in two steps.
         • The second stage involves an autonomous collaborative ascension.
         • The first focuses on identifying the environment and defining the best area for climbing
           • The algorithm uses cloud processing points, unsupervised learning techniques and a relationship proposed by the authors that involves the characterized parameters.
       • Algorithm 1
         • Algorithm 2
   • Experiments and results
     • https://www.youtube.com/watch? app=desktop&v=jYkwVPW-9sY&t
       • Video about the experimental phase is provided, showcasing in detail the synthesis of this section, commencing with the functionalities and characteristics of each robot, the functioning of perception algorithms, and initial indoor and outdoor testing. Various outdoor scenarios with varying heights are tested to verify the system’s functionality and robustness; to enhance versatility, different shots of mission execution are included, both from a first-person perspective (captured by the robot) and lateral and rear perspectives.
   • Main benefits observed from the application of the project
     • Reducing exploration time
       • Increasing areas explored
       • Collaborative interactions
       • Identifying rising zones based on Point cloud
   • Conclusion
     • This article introduces a pioneering approach to tackle the challenge of navigating uneven terrains using collaborative robots, consisting of a mobile ramp robot and an explorer. The method demonstrates an expansion in the exploration area. Collaborative robots emerge as a viable solution for enhancing exploration in rugged terrains by creating an adaptive mobile bridge that links two zones at varying elevations, enabling a second robot to ascend to the unexplored area. The suggested technique for identifying ascent zones via point cloud processing exhibits considerable efficiency, with an average accuracy rate of 95% in zone characterization. This enables seamless exploration through ascent zones determined by the algorithm, facilitating continued exploration into uncharted territories.

Images and files in the map