3D World Building Software Capability for First Responder Robots

Researchers at Sandia National Laboratories have begun developing a '3D World Model Building' software capability for first responder robots. The system illustrates the use of 3D visualization to improve situational awareness in critical situations. It provides users with an immersive, texture-rich 3D model of their environment in minutes using a laptop and color and depth camera. The Sandia software integrates a number of sensors, one being a LIDAR sensor. Sensor data, together with the robot's GPS location and satellite map images from Google Maps, is used to create a working model of the area of investigation.



Transcript

00:00:16 foreign so sandia national labs has recently begun developing a 3d world building system for first responder robots this unique capability illustrates the use of 3d visualization to improve situational awareness in critical situations imagine the following scenario

00:01:10 a response team receives a call to check out a nearby location they've just arrived on site to begin their investigation they launched their robot utilizing the sandia developed 3d world building software as the robot begins its exploration it has no knowledge of the site and the response team is anxiously awaiting a

00:01:32 model of the site to initiate their response the software integrates a number of sensors and cameras mainly a laser imaging detection and ranging or lidar sensor and the prime sense asus sensor which provide both color data and depth measurements the sensors scan the robot's

00:01:55 surroundings and a fusion of this sensor data together with the robot's gps location and satellite map images from google maps is used to create a working model of the area of investigation the model is created through a combination of 2d and 3d images which are layered to produce the most accurate representation of the area possible unknown areas are represented in a

00:02:22 variety of different ways the red dots show each measurement of the live sensor stream provided by the lidar sensor and highlight any obstacles the black lines represent known obstacles the white represents free space and the great areas represent areas that are closed off or unexplored to fill in the unknown areas of the

00:02:51 model the operator drives the robot along various points and takes a number of snapshots since existing robot systems do not record where they have been it's all too easy to forget where any obstacles are located once they are no longer visible a 3d model helps the operator to move beyond tunnel vision and become more situationally aware because they're

00:03:23 moving from a two-dimensional camera view world to a three-dimensional world this software tool can also be used to take measurements within the model to determine the size and volume of any unusual packages or structures in an area models created by one operator can be helpful to other operators as well because once a model has been created it

00:03:54 can be saved and shared with others this software also has the potential for collaborative projects for example a team of robots could be deployed simultaneously to map out a 3d model of an entire compound this 3d model took only 20 minutes to complete the 3d world building system for first responder robots is not intended as a

00:04:22 final product yet we're currently working on improving the visual quality and presentation of the data and we're looking for partners and sponsors to continue work on this extraordinary project you