A growing number of enterprises are automating physical processes and integrating robotics to streamline operations, maximize efficiency, and reduce labor costs. But the robotics industry is still nascent, so there’s immense pressure for project leaders to prove ROI.
On the other hand, if you’re a mechatronics engineer, you might spend years designing, building and refining complicated hardware until your vision is realized. But once your robot is set loose in the wild, it’s hard to maintain visibility of its performance. How do you know the users of your robot are satisfied? Is it doing its job properly or lying in a ditch somewhere?
This sense of disconnect is very different from the world of software. If you built an ecommerce website, you’d have tools baked in to analyze its performance and compare metrics between users around the world. But when you build and sell a robot, your visibility of its performance is often completely obscured - a problem the industry has yet to solve.
Whether you’re a robot user or engineer, there are three key things you want assurance of: Is the robot turned on? Is it doing its job? And is it doing that job well?
It’s time to take the blindfold off
The answer - which might seem obvious - is having an easy way to monitor your robots’ performance from anywhere. But what does this consist of? We can break the core requirements of robot monitoring down into three categories:
- Hardware monitoring. Making sure the physical components of the robot - such as the CPU, memory, hard drive and sensors - are functioning correctly.
- Application monitoring. Making sure the software is doing what it’s designed for and it’s not glitching or causing issues.
- Qualitative monitoring. Gathering detailed, real-time insights about what the robot is doing at any given moment and how effectively it’s performing intended tasks.
All of these aspects of monitoring are essential to gain a robust understanding of whether your robot is effectively fulfilling its duties. Robust monitoring is also crucial for error diagnosis and problem-solving, optimization, and designing future upgrades and enhancements.
What does effective robot monitoring involve?
The Rocos platform is designed to keep you informed about the health and status of your entire robot fleet, in real-time, from anywhere in a world. Here are four ways you can take advantage of the platform’s built-in monitoring capabilities.
1: Use a custom dashboard to monitor live telemetry
One of the biggest challenges when monitoring robots is sifting through the immense amount of telemetry they generate day to day. It’s crucial that you can isolate and surface the specific data you need in any given situation.
The data required can also vary based on who is monitoring the robot. A mechatronics engineer might want to figure out why their robot suddenly powers off at certain times, why it loses its balance, or why certain objects slip through the robotic arm’s grip. In contrast, an operations manager might want to extract data that helps them pinpoint areas for improvement within a wider business automation process.
2: Review past incidents with storage streams
Whether your robots are basic or highly sophisticated, things will occasionally go wrong. And you might not be staring at the dashboard when that critical failure happens. Even if you are watching, it can be hard to diagnose the root cause of a problem in real-time. There may be hundreds of potential causal connections between processes, resources, and the environment your robot is operating in.
For example, imagine your robot experiences a sudden and seemingly random failure in path planning. This failure keeps reoccurring at approximately the same time every day. After a lengthy investigation, you realize that light coming through a building window at a specific time of day is causing overexposure on the robot’s sensors. This means the robot is unable to detect its environment accurately.
This scenario highlights the importance of being able to analyze historical data. Rocos offers the ability to use storage streams - data that’s captured perpetually and stored for later analysis. When a failure occurs, you can go back to that moment in time and dig into the relevant telemetry to figure out what happened. You can also compare historical trends and data streams across different robots and contexts to identify important patterns.
3: Stay informed about critical issues
To effectively monitor a real-life system, you also need to be told when urgent or critical issues happen. That’s where ‘Events’ come in.
In the Rocos platform, an Event is made up of a trigger and an action. The trigger defines a certain aspect of telemetry, and when it crosses a specific value threshold it raises an Event. This leads to an action, which is generally a webhook or API endpoint that calls on the system to notify someone.
For example, your trigger might be a robot’s battery level falling below 20%. When this occurs, the action might be sending an automated SMS or Slack message to the robot’s operator.
Importantly, you can also use the Rocos platform to send a command back to the robot. For example, if the battery level falls below the defined threshold this can trigger a command for the robot to return to its charging station. Or if a robot reaches a certain GPS waypoint, the platform can instruct the robot to remain still, while also sending a notification to a human security guard telling them that the robot needs clearance before entering the next area.
4: See what your robot can see with operations views
Another important monitoring capability that Rocos provides is local and global operations views. At a global level, we use GPS mapping to show where all of your robots are located in real-time. At the local level, we provide a 3D visualization of what your robots can see. This gives your operators more context about the robot’s environment, so they can intervene or adapt missions as needed.
You can achieve this view by importing 3D models of your environment (e.g. a utilities plant, warehouse or construction site) into our platform, and then using live telemetry from the robot to map its position accurately within this digital twin. Or if your robot has a point cloud scanner mounted on it you can render that live data as a realistic 3D environment.
For example, if you run autonomous freight yards you can view the location of all your driverless trucks at any given time, at either the local or global level. You can generate a 3D visualization of each freight yard, making it easier to map routes and optimize the location of containers and storage space. You can also incorporate streams from other sources like CCTV into your dashboard, and use this to monitor things like the safety of human workers on-site.
Robot monitoring is crucial in every industry
Autonomous robots are increasingly sophisticated and capable of incredible things. But a robot that’s not doing its job properly won’t be generating ROI for its business user. It could even bring reputational damage to the company. So it’s critical that you make sure it’s operating correctly.
That’s why monitoring is proving crucial in a wide range of industries. From cleaning robots to site-inspecting drones, driverless trucks to autonomous submarines, Rocos is helping businesses extract the real-time data they need to manage robot fleets effectively.