Autonomous robots are a game-changing technology for enterprises.
After all, there’s no point having the world’s best robot if you need someone with a mechatronics PhD chasing after it 24/7. Robot autonomy reduces the amount of human resources needed to operate fleets, saving time and costs while maximizing efficiency.
But there are misconceptions about the meaning of robot autonomy. Just like ‘artificial intelligence’, the term has been thrown around arbitrarily, to the point where many people are unclear about what truly defines an autonomous robot.
What is robot autonomy?
There are essentially three components that make up robot autonomy. The robot must be able to perceive its environment, make decisions based on this information, and actuate these decisions. It must do all three things to qualify as partially autonomous.
Despite what sci-fi films would have you believe, most robots haven’t reached full autonomy yet - where no human intervention is required in any circumstance. But this isn’t stopping them from having an impact in a wide range of industries.
Today’s autonomous mobile robots (AMRs) are capable of adding significant efficiencies to business processes by performing routine tasks securely and reliably, alleviating pressure on human workers. AMRs use sophisticated mathematical algorithms that can make sense of data from the world around them, enabling them to complete tasks and navigate obstacles.
How does Rocos support robot autonomy?
In the future, robots will be so deeply integrated with core systems and AI that no human intervention will be required. But this dream of full business process autonomy - where fleets of heterogeneous robots perform every task without help - is still a few years away.
At present, all AMRs require human assistance in certain scenarios. This means businesses need a platform to control their robot fleets with precision, without the need to have staff on-site 24/7 or recruit highly specialized operators.
The Rocos platform offers a range of tools to help users control and coordinate their robots, no matter what level of autonomy they’re currently at. This can help you build smarter workflows and gradually minimize the resources required to operate your fleet.
1: Remote teleoperation from any device
Rocos can be easily integrated with your robot’s existing control stack to enable remote teleoperation from anywhere, on any device. You can configure your choice of input device and send commands to robots via laptops, tablets or even game controllers. You can also configure the payload you’re sending, for example to just send variables like speed and direction, rather than sending the whole message - making the API call lighter.
The ability to remotely teleoperate your robots is further enhanced by the platform’s ability to help you ‘see what the robot sees’ through real-time, ultra low-latency video streams. If your robot is equipped with LiDAR or point cloud sensors, you can even use this data to create full 3D visualizations or digital twins of the environment it’s operating in. You can then control robots while their movements are visualized on-screen in real-time.
2: Precision intervention from human operators
Imagine a conveyor belt with a robotic arm that uses cameras and machine learning to recognise and sort different objects into groups. This technology is useful in industries like postal services where they need to sort packages of different sizes. But what if the sorting is only 98% accurate? The 2% error margin could be extremely costly.
Rocos enables real-time and precise intervention from human operators. It allows operators to click anywhere on an image or video feed from a robot, capture the X and Y coordinates, and send these back to the robot with an associated command. This can provide more sophisticated control in a range of situations where the robot cannot achieve full accuracy on its own.
3: Consistent and secure API access
Rocos is designed to make controlling robots as easy and secure as possible. That’s why we offer secure API access to robots in a no-code environment. This means that whoever is developing your Android app, for example, doesn’t need to understand robotics to connect to the robot. They can use our APIs which have been pre-configured by robotics experts.
Through the Rocos portal and API, our customers can set commands which interface directly with the robot’s on-board control software. This allows users to standardize access to their robots in the field, even if those robots are different models or from different manufacturers. Again, no coding is required.
For example, you can have a command called ‘move forward’ and keep this naming consistent across every different type of robot - and it calls the API in the same way. This allows ease of use and consistency of control across heterogeneous robot fleets.
4: Responding to triggers and events
As explained in our article about monitoring your robot fleet, the Rocos platform allows you to define Events - which are made up of a trigger and an action. The trigger defines a certain aspect of telemetry, such as the robot’s battery level. When it crosses a specific threshold it raises an Event, which leads to an action - for example, notifying a human operator through Slack that the robot’s battery level has fallen below 10%.
These Events aren’t just important for monitoring your robot’s performance, they can also be used to send commands back to the robot. For example, if a robot encounters an unexpected obstacle while moving between routine waypoints, the platform can send a message back to the robot. The message could instruct the robot to stand still and take a photo of the object, or to raise an alarm alerting a nearby operator in the facility.
By using these features, you can programme robots to react to conditions in a way that serves your business requirements and supports greater autonomy.
5: Getting up and running
Because of our native support for ROS 1, ROS 2 and MAVLink, it’s extremely fast to get robots up and running with Rocos. It only takes a few minutes to go from installing the Rocos agent on a robot to functional dashboarding. And it doesn’t take much longer to achieve full remote teleoperation with a gamepad controller and video streaming - even with a robot we’ve never encountered before.