Robotics is a field that is advancing at a rapid pace, with newer and more innovative applications being implemented as the technology improves.
In particular, sensing technology has come a long way, and robots are now able to understand the environment around them in much greater ways.
But developing robotic platforms to understand this environment is still a great challenge for engineers.
Mark Bishop, CEO of Ocular Robotics, said as humans, we’re able to direct our senses, such as sight, by turning our heads to look around, and that restricting this would significantly hamper our ability to carry out tasks and how fast we could complete them.
“That was also largely often the case with robotic systems – the ability to redirect their sensing is quite limited, and that affects the speed at which they can operate in an environment and complete their tasks,” Bishop said.
But Ocular Robotics is changing that, with the company using its RobotEye technology platform to enable various types of sensors to be used so robots can make sense of their environment faster and more accurately, carrying out both operations simultaneously.
Early on, Ocular Robotics looked at the different approaches being used to enable sensor directing, particularly on mobile robotic platforms.
“It became clear that really what everyone was doing was just the same thing over and over again and trying to optimise the same old approaches, rather than dealing effectively with the crux of the problem,” Bishop said.
To Bishop, this core problem was trying to move around too much mass, which limited the dynamics with which users could then direct the view of the sensor – while they could change direction precisely, they couldn’t be fast and precise at the same time.
“With the RobotEye technology platform, we’ve allowed most of the mass that would normally be required to be moved to redirect the view of a sensor to remain stationary so that we move just the minimum amount of mass,” he said.
“That’s the key to how we're able to deliver the sorts of very high levels of speed and precision that we do.”
Adaptability
RobotEye technology can be redesigned according to the application. For example, sensor pointing systems can be configured to provide system apertures from as small as 5 mm in diameter to over 200 mm in diameter.
“That core itself can be adapted in many different ways, and then subsequent to that is an optical layer. That optical layer can be adapted to achieve the optical performance that you need,” Bishop said.
“The bandwidth of an optical system for a hyperspectral sensor is much wider than that for a visible band camera, and so all of those optics that we use in our thermal systems are very different from the optics we use in a visible band system.”
RobotEye technology has been used in several different projects.
A UK-based defence and security prime contractor with operations in Australia and the U.S. has used Ocular Robotics’ thermal imaging capabilities because it was the only one that could address the company’s requirements – simultaneously tracking a number of drones that could be launched against a facility or building, for example.
Another application that Bishop said the company had not even considered before was stress analysis for the airframes of military aircraft, such as the wings and pieces of fuselage and testing where and how they might fail.
Personnel at the Defence Science Technology Group, part of the Australian Department of Defence, approached Ocular Robotics to work with them on a system when they saw how it could rapidly and precisely move the view of the sensor to show where stress was.
The system is now in its second phase of development, and Bishop said it could be applied to not only military aircraft, but also police aircraft and across the fatigue testing industry.
RobotEye technology has been possible due to advances in several aspects of technology, including motor and battery technology. Advances in smartphone technology has also meant better resolution and smaller cameras for robots.
Ocular Robotics is working on utilising technological advances to develop other systems that are better and faster.
“A system that we’re developing at the moment will be able to capture full panoramic images of nine gigapixels in less than a minute,” Bishop said.
“That’s precisely due to very high-resolution sensors that are becoming available.”