The vision in industry 4.0
Industry 4.0 or 4th industrial revolution is the convergence between a factory's tools and machines with digital technologies to result in unique and personalized products. It’s a new, more efficient way of producing. After mechanization (Industry 1.0), mass production (Industry 2.0) and automation (Industry 3.0), we arrive at the heart of a flexible system where products are created with a relatively affordable cost, corresponding to the needs of consumers .
​
Each revolution has led to a massive increase in productivity.Industry 4.0 will rely in part on vision technologies because it is an essential element of automated information collection systems: this is how industrial vision systems were created. How do vision systems enable industry to gain efficiency? And how is industry 4.0 revolutionizing the business world?
​
Take the example of collaborative robots which seem to be one of the pillars of industry 4.0. Collaborative robots are increasingly present within companies and their main characteristic is interaction with humans in a common workspace. For example, multiple companies are deploying fleets of autonomous mobile robots in logistics warehouses to optimize productivity. The camera plays an important role in the intelligence of this type of robot: it will capture the image which will generally be processed by a central system.
​
Vision systems and robot intelligence offer a range of capabilities such as identifying and understanding the positioning of objects. The use of vision allows robots to operate as a fleet. Some researchers will even imitate natural functioning to design efficient systems. This is the case of robots inspired by ants which work in swarms to overcome obstacles and lift objects. Industry is the traditional sector in which vision and robotics are found because the environment can be controlled, both in terms of lighting and the layout of the premises.
​
Another example of the use of machine vision concerns the quality controls required during the production process. This is an essential step because it ensures product conformity consistent with the company's level of requirements but also to comply with legal obligations. It is important to consider that quality control is not only found in part of the production chain, but throughout the production chain, through administrative processes of control, visual verification, testing and d 'analysis.
Logistics robots from Locus Robotics operate in fleets and find their way using vision and LiDAR technologies
Example of an industrial production environment where the layout of parts is controlled.
Today, many visual controls are automated via industrial vision systems: dimension control, product labeling, fault detection, etc. Within the controlled environment of production machines and tools, it is possible to set up reliable checks relying on a regular position of the parts to be checked and constant lighting.
​
Regardless of the sector of activity, industrial maintenance is an essential link in the production chain. These are all operations intended to prevent or repair breakdowns that may occur on machines. When a company installs a system to maintain its production equipment, this will allow it to prevent a large number of problems and therefore reduce productivity losses.
Today, augmented reality brings significant added value to companies in the industrial sector and more particularly on maintenance issues. It often refers to the different methods that allow virtual objects to be realistically embedded in a sequence of images. Thanks to this technology, industrial technicians are supported in their operations.
Augmented reality translates as an extension of the intervention environment. Today, the presence of augmented reality in areas such as maintenance, assembly processes or quality control is already common in many companies implementing systems to improve their industrial processes.
Augmented reality, here seen through a mobile application
Today, vision has taken an important place in industrial environments, in multiple ways. However, a need for physical infrastructure (e.g. lighting, stage, etc.) is necessary when implementing such systems, particularly with a view to controlling the observed environment. Two elements also complete the camera, mainly used as an image capture system: a processing center and a communications network to transmit the video data stream.
​
From this observation of the heavy initial investment that the inclusion of vision systems can represent, interest in embedded and autonomous solutions is growing. The use of cameras with on-board intelligence makes it possible to process visual information locally and thus reduce both the volume of data passing through the network and the dependence on the central data processing and analysis system. How does it work ? We will address it soon in an article onembedded intelligence.
