In a contributed article for Internet of Business, Jonathan Wilkins, marketing director at EU Automation, a supplier of obsolete industrial parts, discusses the role of machine vision in the factory of the future.
Birds of prey owe their success as hunters to their superb vision: not only can they see around four to five times further than the average human, but they also see more colours and have a wider scope of vision. It’s what gives them the ability to navigate in the air and identify small prey far below on the ground.
Many industrial organizations could benefit from this type of heightened perception – and here, we see machine vision being introduced to identify improvements and to support intelligent locomotion by robots.
Machine vision explained
In simple terms, machine vision is an image processing technology that enables automated object scanning within a set field of view. Plant operators might mount cameras, for example, on production lines or cells for real-time process control, product inspection and sorting and robot guidance.
The technology enables robots to interpret their visual surroundings, allowing them to move around independently and safely. In other words, they can use visual information to recognize the environment and to make decisions that they are not directly programmed to make.
A camera does not see in the same way as the human eye, but machine vision systems use pattern detection software to examine data and draw conclusions based on prior knowledge.
This technique is particularly useful when inspecting the quality of raw materials and final products for defects. If a problem is found, a part can be redirected or a process corrected to resolve the issue.
As well as flaw detection, machine vision can be used to ensure operations are traceable using identification tags. A camera can read the tags, allowing the information to be used to direct the product or to register which parts are at what stage of the supply chain.
Smart cameras and sensors can digitize and transfer information, decoding what they capture and removing the need for human interpretation. The machine can then decide whether the information needs communicating to a central control system. These are low-cost, easy-to-use systems that can be a good option for those looking to streamline automated manufacturing.
Machine vision is central to the idea of the ‘smart factory’, based upon a communicating network and the intelligent exchange of information among sensors, devices and machines. Acting as the ‘eyes’ of the factory, image processing systems based on industrial cameras can compute information that was previously gathered and analysed by humans. This reduces errors and enables robots to react flexibly to production control needs.
Because image processing equipment captures, gathers and exchanges data, it is a key technology for an interconnected production process. Data can be transmitted to the value chain, but also used to trigger intelligent actions. In an IoT-enabled smart factory, we’ll see more sensors capable not only of understanding images but also identifying patterns and learning from them to inform future decisions and actions.
The technology might be used to examine the state of production machines for wear and tear, for example. This information is useful for maintenance and can alert a plant manager of the need to order a replacement industrial component before it breaks.
Machine vision can also improve plant efficiency through comparison. Visual sensors can record plant operations worldwide and relay the information to data centres. The IoT connects these factories, allowing other plants to compare processes and keep up-to-date with and understand changes in product quality.
With machine vision systems decreasing in size and increasing in speed, accuracy and resolution, the popularity of these systems could grow drastically over the next few years, giving machines the same visual precision as birds of prey.
The post Machine vision: a bird’s-eye view of the smart factory appeared first on Internet of Business.