It's only a matter of time before vision-enabled robots work side by side with humans, according to world-leading researchers at a Queensland university.
The Australian Centre for Robotic Vision has built a machine, dubbed Baxter, with basic hand-eye co-ordination and an ability to process visual stimuli.
That sets him apart from current robots, ACRV director Professor Peter Corke said at the Queensland University of Technology.
"We think this is critically important for robots to move out of factories and into all sorts of other domains," he said.
"Baxter is perhaps the vanguard of this new wave of visually-enabled robots."
But the technology is still in its infancy, Professor Corke said as he sat down to a game of Connect Four against Baxter.
The scoreboard, after two rounds, was tied one-all.
The robot can differentiate between basic colours and perform simple tasks such as picking a ripe capsicum.
"The real world is much more complex than that," Prof Corke said.
"That's the central challenge. How do we take a picture of the world and interpret what's in it."
The practical applications of the technology in the agriculture sector will be trialled in June.
One machine, known as "Ag Bot II", can help scan, fertilise and harvest crops.
The machines can work in a team and communicate with one another.
Researchers also hope to use them to analyse weed samples and interact with manned aircraft.
"The robots can also carry lots of sensors for soil monitoring, but also for pest and diseases, so they can be at the forefront of dealing with invasive species," Robotics Professor Tristan Perez said.
The machine's built-in weather station allows it to make decisions based on conditions.
Other vision-enabled, mobile robots are capable of creating 3D maps of their environment.
One miniature robot can even drive a toy car.
But for anyone feeling a little uneasy about the possibility of a Terminator-like Skynet situation, Prof Corke says robots with a temperament or personality are most likely a decade away.
"Robots just the moment are still a little too simplistic and literal," he said.
But all roboticists dream of creating a machine that can communicate and learn, he said.
"That's the light on the hill. That's what gets me out of bed in the morning."
Share
