Innovations Operator Support
Discover our latest operator support innovations
We developed several innovations that help companies evolve into agile and people-oriented organisations. We let people and machines work together in a smart way. Think about programming robots, using Augmented Reality and improving the ergonomics of an industrial work cell.
Other themes:
Versatile, mobile third-hand operator support
Operator support in industry is becoming increasingly important. Cobots can be a great solution for heavy, repetitive tasks. They work continuously and you can use them in a different way every time. Moreover, Cobots also ensure that operators stay motivated by freeing them from certain boring, heavy routine tasks.
To support operators, we developed a mobile cobot. The cobot helps the operator to move heavy objects, and immediately holds them in the right place for assembly. This way, the operator can perform the assembly in a smooth way. Moreover, the cobot is compact, especially compared to, for instance, an overhead crane. In addition, the cobot is more mobile, and more versatile since you can program it for different tasks.
Dynamic operator profile classification for adaptive HIM DWI platform
Assembly operators have different levels of skill and experience for certain tasks. Adapting the content of the work instructions to the skill and experience level of the operator ensures that the operator will perform better. To achieve this, Arkite added an adaptivity module to its HIM platform. This module ensures that operators have to identify themselves before they can use the system. Unfortunately, privacy legislation often throws a spanner in the works for this type of identification solution.
Therefore, at Flanders Make, we have developed a tool that automatically and dynamically detects the skill level of the operator for each task he/she performs. The tool then adapts the work instructions. For this, we only use the observations made by the Arkite Human Interface Mate (HIM). Without any other form of identification, our profile classification system achieves an accuracy of 95% compared to the current classification method used by Arkite. The great advantage is that the operator does not have to identify himself, but still receives adapted work instructions.
Improving the ergonomics of an industrial work cell
Not being able to work has an enormous impact on the economy. Every year it costs Flanders billions of euros. Muscle disorders, which are caused by working in non-ergonomic conditions, are a major cause of operator absenteeism in industry.
Optimisation of both work cell and operator's mindset are necessary to minimise injuries caused by poor ergonomics and to improve working conditions. In this demo we show how to use a model to predict and improve the ergonomics of an industrial work cell. By using intuitive visualisations we help operators to recognise non-ergonomic postures during their work. Over time, we monitor the ergonomic load on individual parts of their body and encourage them to improve their posture through gamification and personal challenges. This not only leads to improved (mental) health of your operators, but also to less work-related absenteeism.
ErgoEyeHand
Not being able to work has an enormous impact on the economy. Every year it costs Flanders billions of euros. Muscle disorders, caused by working in non-ergonomic conditions, are a major cause of absenteeism of operators in the industry.
To improve the working conditions of your operators, a cobot is a good solution. A cobot can take over tasks that are heavy, dangerous, precise or monotonous. With the help of cobots, operators can improve their work posture, and thus reduce the risk of ergonomic strain. Results show an improvement of up to 50% for one of the most important ergonomic factors - REBA (Rapid Entire Body Assessment). This means that the use of cobots not only improves the (mental) health of your operators, but also the absenteeism related to work.
Robot programming: Learning by XR Demonstration
The precise programming of the robots and cobots in your production line is not only a difficult and time-consuming task. It also requires specific expertise. To facilitate this process, we developed an "Extended Reality" framework that stores the movements of operators and translates them into a robot trajectory. After a short training, an operator can adjust and simulate trajectories before the robot executes them. By avoiding conventional online programming, we can reduce the downtime of your production line by up to 95%. Moreover, we use the operator's knowledge as input, which allows us to improve the XR simulations and reduce the total programming time of robots by up to 50%.
Modular toolchain for creating and executing digital work instructions with Augmented Reality
Augmented Reality (AR) is a promising technology that can facilitate production processes. However, its use in an industrial setting remains rather limited. The reason is that developing AR applications is a time-consuming task and, moreover, work for experts. This often leads to customised solutions that are expensive and difficult to deploy in low-volume and high-variability production.
To address this issue, we developed an authoring tool that demonstrates how simple AR instructions can be generated from a CAD file, without programming and with a simple user interface. This allows operators or method engineers to create their own AR work instructions without being distracted by programming. The full focus can thus be on the creation of the work instructions themselves. Moreover, the generated instructions can be viewed on any device such as projection systems, tablets and AR glasses. Depending on the complexity of the assembly and the detail of the AR instructions, it takes +/- 1 working day to get instructions for industrial applications up and running. This is a huge time saver compared to custom programming a particular application. By using this solution you can support your operators in a fast way to reduce lead times.
Teaching robot force control by demonstration with a haptic device
Many assembly processes in the industry consist of delicate and precise tasks. Human intervention plays an important role. This makes these processes expensive, so they are moved to cheaper production countries. By robotising parts of a production process, we can make the process cheaper without compromising on quality. But the control process of a robotic gripper is complex. Moreover, existing state-of-the-art solutions are limited to just gripping and holding, without any other advanced actions. Therefore, currently, one usually uses complex and expensive multi-finger exoskeletons for this work.
At Flanders Make, we have developed a tactile feedback device with a data-driven control that makes it possible to grasp and manipulate objects from a distance using a robotic gripper. In addition, we developed a learning framework where our setup automatically reproduces your interaction with an object. This technique allows us to keep our assembly processes here for delicate and precise tasks without increasing costs or sacrificing quality.