How can usage data teach people to operate machines?

There are always two sides to operating an industrial machine: the abilities of the machine and the operator’s ability to use the machine. Both of these sides are important when looking for improvements in performance and productivity, but sometimes the human performance side is forgotten, says Matti Lehto, Director, Product and Engineering Process at Konecranes. By keeping track of the operator’s performance, sensor technology and usage data can teach the operator to perform better. But are we ready to be analyzed and evaluated?

Let me begin with an example in modern cars: the car can tell you your average fuel consumption per 100 kilometers by the deciliter. Car motors are being heavily researched and developed so that they would use even one deciliter less fuel. At the same time, the way you drive affects the fuel consumption to a great extent. What if your driving was monitored, and the car could tell you if you make a systematic mistake that brings the fuel consumption up?

For example, every time you stop at a red light, you use the brake with the clutch pushed in when you could slow down the car by engine braking. Changing this habit could bring down the fuel consumption in city driving by an entire liter. Would it therefore be more useful to educate the driver to become a fuel conscious driver, because there is a bigger potential for fuel saving?

With the help of new software, the sensor technology would already allow this. The software could calculate the optimal fuel consumption for a certain route, compare it with the actual fuel consumption and analyze why the actual consumption was more than the optimal consumption.

In the case of industrial machines, most companies don’t have a “driver’s license” or another way of saying when a person has the needed skills to operate an industrial machine. Nevertheless, the skills of the operator affect the overall productivity. With industrial internet and the help of sensors it is possible to measure and analyze how smooth the operation of a crane is.

For example, it is possible to place sensors in a crane to measure how much the load is swaying and how much time the operator uses to stop the swaying – by waiting for the swaying to end or by moving the crane to make the swaying stop. By collecting usage data with sensors, it is possible to gather information of the performance of a single user to evaluate if his or her performance is already on a good level or if it should still be improved. If the operator needs more practice in operating the crane, they can receive further training to operate the machine.

Following the example with modern cars, if the car could monitor itself and the driving data could be transferred to a data center for analysis, and the driver would get a regular report about his or her driving performance, this would help the driver to become a better driver. This type of service could be very useful in operating industrial machines, too. It would be hard to imagine this kind of a feedback system to be created with any other solutions than the technologies provided by industrial internet.

Are we ready to face the facts?

Closely following the performance of the operator also improves safety. One single accident in a factory can easily cost you tens or even hundreds of thousands of euros. Therefore investing in a service that can prevent accidents can pay itself back surprisingly quickly.

Airplanes have black boxes that store information about the operation of the plane. In industrial environments, cranes, forklift trucks and other moving machines could also have ”black boxes” that keep in its memory all the movement and actions of the machine from the last few minutes. If something unexpected happens, an accident or even a close call, you have an accurate snapshot of the last minutes. This can help you avoid the same mistake in the future.

The best tool for better safety is accurate fact based data like this. For example, in industrial environments, this fact based data could prove that driving forklift trucks at high speed can lead to more accidents: when you have cold hard data to prove your point, it is easier to justify speed regulations. If it is only a gut feeling, people will not react.

The technical solutions are already available. The bottleneck of having these things in operation is that the industrial companies – and also the car drivers in the case of cars – would have to be ready to give out this information about their machine and their own behavior for analysis. This requires a certain attitude. You have to believe that this information can add value and make your performance better. The question is if and when we are ready for this.