Previewed Reality: System foretell shifts in their surrounding


When robots and humans interact in a shared environment, they need to move in ways that prevent collisions or accidents. To reduce the risk of accident roboticists have developed various methods that monitor the environment, identify safe track for a robot, predict the future actions of humans moving, and control its movements accordingly.

Predicting human behavior and human user’s movements can be especially challenging. Determining the future movements of a robot could far easier, as robots are generally programmed to perform particular actions or complete specific goals. If a human user can predict the movements of a robot and the effects of the surrounding, then he/she should be able to adapt his/her actions to avoid accidents or collisions.

Researchers at Kyushu University Japan created a system that allows human users to anticipate future changes in their environment, which could guide their actions and inform their decisions. Details of this system were presented in a paper published in Advanced Robotics, compiles a dataset containing information about the position of furniture, objects, humans, and robots within the same environment then produces simulations of events that could take place in the near future. These simulations are given to human users via Virtual Reality (VR) or Augmented Reality (AR) headsets.

To collect information about the position of different objects, humans, and robots in a shared environment, the researchers used strategically placed sensors that include an RGB-D camera and optical trackers. The optical trackers monitored the movements of objects or robots, and the RGB-D camera track human movements. The data collected by the sensors was fed to a motion planner and a dynamics simulator which helps the researchers to forecast changes in a given environment and integrate images of events that are expected to happen in the near future from the viewpoint of a particular human. Human users could view these synthesized images simply by wearing a Virtual Reality headset or an Augmented Reality display. On the headset or display, the images were laid over those of the user’s actual surroundings so that it clearly shows the changes that could take place in the future. Previewed Reality assists as an alternative and as a supplement to more conventional collision avoidance systems, that are designed to identify safe track for robots. The system could be used to improve the safety of interactions between humans and robots in a variety of indoor environments in the future. 

The researchers plan to expand and simplify the system they developed, for example by creating a smaller and more affordable version that can be obtained on smartphones or other portable devices.