The Digital Twin can improve production and reduce downtime – Data gathered into digital twin technology can be used to optimize equipment, says Mathieu Bélanger-Barrette.
A digital twin is a virtual replica of a physical asset. But it’s a lot more than that. The digital twin is a combination of IoT, machine learning, and AI put together to create a replica of your physical machine. It will analyse and predict the impact of actions on that machine.
A digital twin draws on the Big Data trend. We have entered an era where data is one of the most important resources companies can get, either for their users and their machines. The goal now is to gather as much data as possible in order to detect trends or to react to given situations.
Moving to Predictive Solutions with the Digital Twin
If you’re managing a robot, you’re observing trends. If you see a particular use of a certain robot, it may prompt you to perform preventive maintenance more frequently to keep the unit from going down. What we generally see in the industry is reactive and proactive solutions to problems. With a digital twin – once you have enough data, experience, and a sufficient sample size – you get the ability to move to predictive, and eventually, prescriptive solutions.
The digital twin is not simply a robot or a turbine that acts like the actual machine that you are operating. The digital twin the summation of all the past experiences and previous data about your physical machine joined into a digital representation. The digital twin can draw on part experiences to predict when a certain failure or other unwanted event will occur, and it can learn how to avoid that event.
Good Principle, But How Will a Digital Twin Help Me?
Car owners know very little about their tires, brakes, engine, or transmission before they get to the garage. Imagine that a digital twin of your car could give you reports and recommend maintenance based on its utilization. Tesla has all its cars connected to the cloud. They can likely log data from all sensors placed in all cars. They are putting this together in order to build an accurate representation of their cars and predict what will happen if one performs a certain action. This goes beyond maintenance and enters into deeper decisions that the car can make.
According to the weather forecast, the wear of your brakes, and the banking of the next curve, there is a reduced chance that you can take that curve at a given speed without using traction control. In the future, given all the gathered data, autopilot would be able to bring the speed down a notch to prevent a potential crash.
Saving Money with the Robot’s Digital Twin
Gathering data on equipment or products is an investment. If you can use that data to optimize your cycle time, reduce downtime, or improve your production, the investment may be returned quickly. We’re headed toward the time when your robot will be able to optimize itself or make decisions based on previous experience. We will eventually get to the point where the robot will ask you to optimize its paths or seek permission to perform a certain action in a particular situation.
For now, keep logging your data. Who knows what you will be able to do with it in the future. As a production engineer, I recommend that you to start now to capture the heartbeat of your production. You will then be able to determine trends and eventually you’ll be able to take action to either optimize your production or to avoid breakdowns.
Mathieu Bélanger-Barrette is a production engineer at Robotiq, where he strives to constantly optimise the production line for Robotiq grippers.