Scientists all over the world are developing intelligent robots that can observe and learn from experience and human behaviour. The robots can then collaborate with humans across a range of sections.
As technology advances, the range of potential applications for robots grows even wider. Robots are becoming more flexible, accurate, safer and faster. These are more energy efficient, compact and lighter. Modern robots are easier to program and capable of integrating with a wider range of machines.
Use of industrial robots has increased steadily in recent years in India. Along with with the cloud, these play a key role in production, as the most flexible elements. Robotic automation is indispensable for anyone wishing to rise to the challenges of Industry 4.0.
Amit Maheshwari, technical lead – robotics and automation, Ingen Dynamics Inc., says, “Latest technology in robotics is integrating robots with artificial intelligence (AI), so that these can interact with humans in a better way. This is done by making the robots understand emotions, faces and expressions of people. The robots could then replicate the same, and interact in ways similar to having a normal conversation. We would not know over a phone call if the other side is a human or robot.
“Personal robots is a recent application area. Warehouse automation is another. Robots like unmanned vehicles are being used in the defence sector. Exploration is going on with underwater robots. Next stage of evolution is more flexible industrial robots.”
Achu Wilson, co-founder, Sastra Robotics, explains, “What separates smart robots from traditional industrial automation is the capability of the robots to adapt to various situations and work as intended. This capability is mainly brought in by advances in AI, machine learning and sensors, among others.
“Traditional robots work excellently in well-defined, controlled factory or production environments. These do not have sensing or decision-making capabilities to operate in regular, cluttered human environments. For example, robot manipulators are used in factories to pick/place/process thousands of similar items. But these cannot perform well with other objects.
“Amazon’s picking challenge aims to develop such intelligent capabilities for robot arms that enable these to pick up thousands of varying objects in a typical e-commerce warehouse. These sensing and decision-making capabilities to operate in unstructured human environments will make the robots smart.”
Cobots for industrial applications
A collaborative robot, or cobot, works interactively with humans in a shared workspace. It helps the human operator(s) to fulfill tasks and minimise risk in situations such as transportation and handling of sharp, pointed, hot or heavy work pieces.
Cobots are easy to use, flexible and safe. These are easy to program and can learn independently using machine learning. Instructions can also be given using a graphic user interface other than coding. Anyone can flexibly program cobots for a variety of tasks.
Wilson adds, “Application areas of smart robots are mostly in human environments. These days, cobots are increasingly used in production environments, because these can augment the capabilities of humans to work safely hand-in-hand with them.
“Another big application area is consumer robotics, which includes robots that can work in human homes.”
Cobots versus industrial robots. Cobots are designed to work along with human operators, while industrial ones work in place of humans. Cobots are capable of self-learning during a job, while the latter require an engineer to write new code for any change in process.
Cobots are not designed for heavy manufacturing as these work closely with humans. Hence, these are safe enough to function around humans. Industrial robots, on the other hand, can handle heavier and larger materials, and require fencing to keep humans out of the workspace. Cobots immobilise at the slightest touch due to sophisticated sensors and, thus, prevent any danger to nearby people.
Developing smart robots
Building smart robots requires automation, AI and machine learning, for human-robot collaboration. Integrated sensors in robots have made it possible to handle assembly tasks accurately.
A smart robot detects its environment, learns from it and responds accordingly. To detect the environment, it requires sensors like lidar, temperature, depth, proximity and camera. The sensors interact with the environment in real time and generate the required information and responses. The robot checks the information using various algorithms to generate the required responses as per the situation or scenario. It then decides how to act.
Pradeep Shoran, assistant general manager – marketing, Kuka Robotics, says, “We need to understand the requirements of a customer and then build our solution around it. The trend is moving towards greater customisation, more product variants, away from rigid mass production. Hence, personalised products that a customer can self-configure on a computer and order on the Internet are in demand.