Artificial intelligence is transforming the way we live and work.
The manufacturing sector is the one that expects the most from the productivity gains generated by computer vision. This is indicated first and foremost by the wide range of applications in this sector. These range from operational use in automated precision cutting and welding, to processes such as quality control; but also inspection, monitoring and packaging.
In general, computer vision can speed up production without compromising quality, but rather increasing and standardizing it, which is why we talk about vision systems for robotic factory automation.
How machine vision is applied to robotic factory automation
Machine vision is used in robotic manufacturing, where the robot is guided by the camera to move and process or manipulate the product. Through computer vision devices, robots are able to analyse, see and, above all, interpret their surroundings, greatly enhancing their ability to work completely autonomously according to a recipe.
In the manufacturing sector, artificial intelligence has already enabled robots to weld with precision and learn to detect microscopic defects faster and more accurately than any human, also improving production line maintenance.
Robot guidance applications from machine vision
According to the Frankfurt IFR (International Federation of Robotics), the main goal of using artificial intelligence techniques in robotics is ‘to enable machines to adapt, in real time and with increasing ease, to the variability and unpredictability of the environments in which they operate.’
In the field of robot design and programming, there are two types of application: one dedicated to the acquisition of skills that enable the machine to identify its environment and respond to it autonomously and in real time, and the optimisation of processes, to improve, for example, the robot’s control skills.
With respect to the second application, we find all the studies and vision systems for quality control, which, for example, allow for inspections and controls beyond human capabilities.
While for the first application, i.e. the tasks of sensing the external environment, they require sensors and cameras installed on the robot itself to help it locate itself and the objects it has to manipulate. Furthermore, the robot’s real-time response activities rely on machine learning algorithms that allow in-depth analysis to choose the most appropriate course of action, such as the right amount of force in grasping a tool or the right position of a gripper in picking up an object, or ultimately the correct path to take to position it in a given spot.
Benefits of machine vision systems for robots
Artificial vision systems for robots therefore bring benefits especially in the presence of unpredictable situations, helping to limit the increased level of variability in the environment. For example, when the objects to be picked are, for the robot, of different types, mixed together, not sorted in the containers, or where the task involves the robot loading the shelves differently each time.
Computer vision systems also speaker where it is necessary to ‘correctly identify the contours of similar, compact objects, such as packages of the same colour or textured surfaces such as wood’.
Thanks to artificial intelligence planning and programming software, designing the most efficient motion path for a robotic arm takes a fraction of the time it takes an engineer to manually programme the same path. This can, for example, turn a 90-minute maintenance task into a 2-second adjustment.
‘Smarter, more efficient robots also help meet sustainability needs through energy efficiency, waste reduction and a smaller operational footprint.
For example, automotive precision coating technology can reduce material waste by up to 60 per cent and improve sustainability by reducing energy and water use.
In the food and beverage packaging sector, intelligent automation allows manufacturers to adapt to recycled packaging and reduce the use of plastic.’ Argues Marina Bill, president of IFR.
The use of robot guidance in robot lines therefore allows for increased speed, flexibility, efficiency and versatility in the plant.
Artificial intelligence allows robots to be more capable of learning, able to learn through experience rather than programming, and able to work in dynamic environments or in the presence of people.
Robot guidance, AI and the future
We have discussed the application trends in robotics in a previous article, let us now look at what are the trends in current studies concerning robotics and artificial intelligence.
The research working on the application of computer vision in robotics is currently working on perfecting the machine’s understanding of voice commands, and on semantic intelligence, i.e. the ability to enable the robotic system to understand the context and the person it is interacting with and consequently make appropriate decisions.
Robots and artificial intelligence thus offer companies the flexibility they need to respond to rapid changes in demand and to remain competitive and look to the future.
For a Kablator technology consultation:
+39 3313466711- info@kablator.com
