Demonstration Videos of PROWL: Perception for Robotic Operation over Widespread Lighting

The following videos exhibit work accomplished by TRACLabs for the PROWL project, "Perception for Robotic Operation over Widespread Lighting".
Click on an image thumbnail to view an entire video (m4v format). Or download the ZIP file (230 MB) that contains all videos (other than the summary).

14 MB
Narrated Summary Video
This video summarizes the system seen in the longer video with narration.

73 MB
Demo - Indoors
The PROWL system mounted on the robot, TRACBot, using its pan-tilt arm and base to track and follow a user according to user gesture inputs. This demo was performed indoors in a well lit room.

The robot in this task is fully autonomous. The user is holding the robot's e-stop.

44 MB
Demo - Outdoors
The PROWL system mounted on the robot, TRACBot, using its pan-tilt arm and base to track and follow a user according to user gesture inputs. This demo was performed outdoors on a warm, sunny day in a parking lot.

The robot in this task is fully autonomous. The user is holding the robot's e-stop.

18 MB
Demo - Dark Room
The PROWL system mounted on the robot, TRACBot, using its pan-tilt arm and base to track and follow a user according to user gesture inputs. This demo was performed indoors in a dark room. Since the making of this video, the 2D image's enhance feature no longer fully saturates to green and the RGB pointcloud no longer freezes when complete darkness results in no points in the RGB point cloud.

The robot in this task is fully autonomous.

38 MB
Demo - Night Time
The PROWL system mounted on the robot, TRACBot, using its pan-tilt arm and base to track and follow a user according to user gesture inputs. This demo was performed outdoors at night in a parking lot.

The robot in this task is fully autonomous. The user is holding the robot's e-stop.

3.2 MB
Demo - Smoke
The PROWL system tracking (in thermal 3D) a user in a warm, smoke-filled, well-lit room. Although the smoke occludes the vision of RGB cameras and affects the RGB point cloud, the tracking and gesture recognition using thermal cameras is unaffected by the smoke.

21 MB
Demo - Web Interface and Painting RGB Cloud with Thermal Data
The web interface for the PROWL system, showing various display options including painting the RGB cloud with thermal data. This demo was performed indoors in a well lit room.

11 MB
Demo - Obstacle Detection Indoors using RGB or Thermal
Visualizing the PROWL system's ability to detect obstacles using RGB or thermal point clouds. This demo was performed indoors in a well lit room. Illustrates the limitations of the thermal point cloud at obstacle detection in indoor, air-conditioned environments.

The robot in this task is being moved manually.

26 MB
Demo - Obstacle Detection Outdoors using RGB or Thermal
Visualizing the PROWL system's ability to detect obstacles using RGB or thermal point clouds. This demo was performed outdoors on a very hot and sunny day. Illustrates that the thermal cloud can likely be used for obstacle detection in such environments.

The robot in this task is being moved manually.


Updated: July 22, 2015