The efficiency and productivity benefits of robotic manipulation allow for the expansion of its application scope across factories, medical, pharmaceutical, biological and operational services.

New research from global growth consultancy Frost & Sullivan finds a renewed interest in robotic manipulation, especially with the increasing uncertainties and complexities of the real world. In manipulation, an important robotics research area, the focus is now slowly moving toward developing remote manipulators. Present day robotic manipulation research addresses issues such as mobile manipulation in poorly modelled environments, learning, adaptation, the development of real-world manipulation, emulation of human-like behaviour, human-robot interaction as well as cooperation, and design of forced controlled and compliant manipulators.

The industry also continues to work toward improving robotic manipulator end effectors dexterity, and minimising or reducing the number of configuration parameters without compromising precision positioning with distributed manipulation.

"Perception-based learning for robotic manipulation tasks involving the use of force/torque sensors represents a hot area wherein qualitative spatial reasoning is done from a bottom-up perspective," notes Frost and Sullivan research analyst Vishnu Sivadevan. "Much work has also been done on intelligent robotic motion using sensory input for flexible manufacturing cells."

The world's first magnetic resonance imaging (MRI)-compatible surgical robot, 'NeuroArm,' represents a notable advancement in the field of robotic manipulation. Performing a surgery requires enhanced spatial resolution to view the parts at a cellular level rather than just the organ level. The newly developed surgical robot meets these enhanced spatial resolution requirements and could also assist with operative medicine as well as surgeries.

Despite such advances, there still remain significant technological challenges related to improving manipulation skills such as hand-eye coordination, transport, alignment and grasping. "The robotic visual servoing system, which refers to the use of visual feedback for coordinated movements of a robotic arm, uses only image-based two-dimensional feature tracking," says Sivadevan. "On the other hand, the human system takes into account the three-dimensional features of the environment. Thus, the visually guided grasp mechanism exhibited in humans requires observance and understanding of three-dimensional and not just two-dimensional geometric features."

Going forward, powerful computers will control the next generation of robots. A number of promising technologies such as neural networks, artificial intelligence, fuzzy reasoning, and many other techniques are expecting to find applicability.

Thus, research should focus on developing algorithms and automatic planners for robotic manipulators. Moreover, research on articulated arm design, end effectors design and kinematics, as well as behaviour-based design of robot effectors needs further study and funding.