Get Video-Ready for the Fourth Industrial Revolution
by Puneet Gupta (Senior Manager, System Solutions)
A recent report from Capgemini indicates that smart factories will add $1,500 billion to the global economy by the year 2022. All this rides on the wave of Industry 4.0 – the next industrial revolution fueled by networked and automated industrial devices and systems. The term Industrial IoT (IIoT) quite aptly describes the core of this revolution – integrating IoT, AI, big data analysis and robotics as part of a bigger vision to enhance productivity. While the Americas and Western Europe are leading the charge for industrial automation, the rest of the world is catching up.
Machine vision will have a big role to play in the evolving digital ecosystem. According to McKinsey Global Institute, IoT applications will amount to $3.9-$11.1 trillion by 2025, and this could be fueled by video analytics in a big way. Smart cameras, video-enabled robots, intelligent surveillance systems, self-driven ground/aerial vehicles and smart visualization solutions are some examples to the point. In a way, many products/solutions in the market are already leveraging machine vision, deep learning and visual analytics. The key applications are:
What industries can look forward from these solutions is a heightened level of dependency on machines to do repetitive tasks or chores that are today managed by humans. While some may argue that this will cut into employment rates, what actually is more likely to happen is that human workforce will engage at the next level – monitoring, analyzing and deciding on ways to improve productivity and efficiency of operations.
A self-navigating robot can record the number of obstacles on its route from point A (where it picks an object) to point B (where it places the object). This data, gathered across all such robots, can help the floor manager decide how he/she should stack organize the factory floor for best efficiency, reducing the average time these robots need to maneuver the obstacle.
The other benefit of automating some of these applications is improved success rate or accuracy of performed tasks. Unlike humans, machines do not have issues of fatigue, concentration lapses and ill health. Therefore, using intelligent vision-assisted robots for chores such as packing, counting, sorting, binning and organizing manufactured goods yields higher overall effectiveness. And it avoids the need for a second tier of personnel to do quality checks. Fewer errors translate to far fewer rejected orders, recalled goods and customer complaints.
Productivity is affected in another positive way using machine vision for IIoT. And that is through access to valuable insights and data that were hitherto limited to the ‘gut feeling’ of someone in charge or governed by the keenness of observation of employees. With these additional pieces of information, decision makers can save valuable time.
In a warehouse where inventory management is critical, camera-fitted bots can be used to survey the stacks of goods and identify racks where more items can be stored. Similarly, smart cameras can be used in shipping containers or packages to assess if more goods/items can be fitted in using visual analysis rather than someone explicitly trying to fit them in, particularly for odd shaped items such as furniture or sculptures or toys.
What’s more, with visual analysis, the act of managing building/plant operations becomes secure and predictable! Authentication using biometrics or face recognition not only prevents unauthorized access but also automatically documents the movement of personnel and their activity logs. And it does so automatically, discreetly and in a manner that is foolproof for evidence management. Guaranteeing such a safe environment is a key priority for many industries.
The team at Ittiam is dedicated to finding innovative solutions for enabling these and more applications in industrial automation. Check out adroitVista SDK, a suite of software solutions for machine vision, deep learning and visual analytics that leverages DSP/GPU acceleration to realize real-time visual analytics.