📜 ⬆️ ⬇️

Highly accurate coordinates (+ -2cm) for virtual reality, kopter and robots

Good afternoon, colleagues.

One of the major problems for virtual reality systems is determining the position of a person in it. If modern gyroscopes and accelerometers calmly cope with corners and turns, then everything is much more complicated with a position in space.

Similar problems are experienced by indoor copters. They can hang without turning. Even the height above the floor can be kept at the expense of altimeters. But they float horizontally, because the coters do not know their coordinates and cannot independently adjust their position in space. Inertial systems do not allow this.
')
But there is a solution - the use of high-precision indoor navigation system.
Slightly more details on the system itself: Indoor "GPS" with an accuracy of -2 cm .

Demo:

Demo: Tracking the position of the helmet of virtual reality indoors with a frequency of up to 16 Hz and an accuracy of + -2 cm:



The helmet itself with an installed mobile lighthouse near:




Demo: Walking in a virtual reality helmet:



Demo: Tracking the position of the copter indoors:
- A mobile lighthouse weighing 25 grams is installed on the copter
- 4 stationary beacons hanging on the walls



Another demo copter. We don’t really know how to fly, but we track the position for sure :-)



Demo of a small robot for cargo delivery at assembly plants:
- The robot moves between the “three workplaces” completely autonomously and transports cargo weighing up to 2 kg. The demo shows how a robot can be used in assembly plants, warehouses, factories
- Blue dots - coordinates of the robot, measured in the navigation system. The robot receives them from the navigation system and uses it to adjust its movement.
- Yellow points - coordinates of the robot in its own inertial coordinate system
- Yellow thin lines on the map - three-point path for the robot. Points can be arbitrarily many and the path can be arbitrarily difficult. And his robot will pass with an accuracy of several centimeters, carrying out the tasks assigned to him



Tracking multiple objects at once:
- Since the system uses time division multiple access, the speed of updating the coordinates for each monitored object decreases in proportion to the number of objects. However, in many cases, especially for large objects with large inertia and slow objects, this speed is enough:

Demo: Walking in the eight with one mobile beacon



Demo: Boating with two mobile lighthouses:




Demo: Walking with four mobile lighthouses:



Conclusion

Knowing the coordinates of people, copters, robots with high accuracy and in real time, one can build fundamentally more serious virtual reality systems than exist now, and create copters and robots capable of performing practical tasks indoors indoors truly autonomously.

Source: https://habr.com/ru/post/275759/


All Articles