DAQRI Smart Helmet™ features a revolutionary Visual Inertial Odometry (VIO). This system is part of our Intellitrack™ computer vision suite, which also includes 2D and 3D recognition and tracking, image matching, and more.
So, what is Visual Inertial Odometry?
It’s a clever and unique way to have a machine, namely DAQRI Smart Helmet, know where it is and where it has been. If we explore the components of VIO, the “Visual” part refers to the on-board cameras which make sense of the world, not unlike how your own eyes do. The “Inertial” part refers to sensors which include micro-electromechanical systems (MEMS) devices such as accelerometers and gyroscopes. These let the helmet know how much you are moving and which way is up. And lastly, the “Odometry” part refers to how far one has traveled. It is similar to the display in your vehicle known as an odometer. DAQRI Smart Helmet accomplishes its odometry by performing software algorithms on the data it collects from the visual and inertial sensors.
Putting together the hardware that gives the visual, and inertial data, with the software that performs the odometry, creates a very powerful technology suite that provides DAQRI Smart Helmet with the knowledge of where it is in the world. Using our understanding of physics and math, it can figure out the path you took through a factory. So now you might be wondering — how exactly does it determine where you are and where you’ve been?
Our computer vision uses images of the world around it to help figure this out. It takes a series of snapshots and changes them to black and white to find all the edges in that world. Then, it finds the regions where two edges meet and records these corner points. The software does this for every snapshot it finds and keeps track of the points which seem to appear in every image, but move a little bit. The software also uses the inertial sensors to understand which way is up and uses math to determine how fast it is moving and in which direction. From all this data, it understands location frame-by-frame and records this data, giving the user the ability to track their own path through a factory or work site.
Where can Visual Inertial Odometry be used?
Unlike with GPS signals, we can map out where people are in buildings or even deep underground, like in mines. When you are unable to reach a GPS satellite, Visual Inertial Odometry can take over and let you know where you are and where you have been. This means that the technology can be used in disaster sites, like earthquakes, where rubble and debris interfere with GPS signals, helping first-responders find those in need faster. This same technology can also be used on other technology, such as bipedal or quadruped robots and even self-driving vehicles.
Why is Visual Inertial Odometry important?
Visual Inertial Odometry is important because knowing the position of a piece of equipment or the entire layout of a factory is a very basic need in industry. Combined with augmented reality technology, one can use this data to place augmented work instructions without the need for markers, placards, or posters, to simulate the placement of new equipment, and even to create 3D maps of new environments very quickly. Empowering workers through this technology begins the next revolution in industry, where workers are connected to their equipment in bold new ways, improving productivity and, more importantly, their overall safety.