Robotic Hands: First-Time Experience & Control | Sensors & Beginner Use

Developing robotic hands capable of human-like dexterity remains a significant challenge for robotics firms, despite decades of research and development. While progress is being made, creating a robotic hand that can reliably and efficiently manipulate objects in the real world continues to push the boundaries of engineering and artificial intelligence.

Shadow Robot, a UK-based company, exemplifies the ongoing effort. Founded in the late 1990s, the company initially built robotic hands using readily available materials like wood, springs, and rubber bands, as Rich Walker, a company director, recalls. Today, Shadow Robot’s hands utilize cylindrical “forearms” housing small electric motors – actuators – that precisely control metal tendons mimicking human finger movements. These hands aren’t intended for immediate deployment in consumer products, but rather as a crucial development kit for researchers.

“What we have is essentially a development kit for dexterity,” Walker explained. “You get this hardware, you explore what can be done in terms of dexterity, then that helps you work out what you want to build if you’re going to build a bigger system, or a bigger project, or deploy something.” Around 200 of these hands are currently in use by universities and tech firms, serving as a platform for experimentation and innovation.

The difficulty in replicating the human hand stems from its inherent complexity. Bren Pierce, founder of robotics start-up Kinisi, succinctly puts it: “I think the hand is the hardest, most complex part of any humanoid robot.” Kinisi’s KR1 robots, currently undergoing trials in commercial settings, demonstrate this challenge by offering interchangeable grippers – from strong “gorilla” pincers for heavy lifting to delicate suction devices for fragile items. The need for specialized grippers highlights the limitations of a single, universal robotic hand design.

The demand for advanced robotic hands isn’t merely academic. The vast majority of tools and devices are designed for the human hand, meaning robots operating in human environments require comparable dexterity to interact effectively. This is particularly relevant as robotics expands into areas like manufacturing, logistics, and even healthcare.

The control mechanisms for these robotic hands are also evolving. As demonstrated in a recent project highlighted by Instructables, interactive control is achievable through the use of gloves and Arduino microcontrollers. This approach allows for a more intuitive and direct mapping of human movements to robotic actions, opening possibilities for remote operation and telepresence applications.

However, even with advancements in control systems, achieving natural and fluid movement remains a hurdle. A recent experience with Shadow Robot’s hands illustrates this point. When a reporter was given simple instructions and fitted with sensors on their fingers, they were surprised by the smoothness and precision of the movements, successfully manipulating blocks and cups. This suggests that significant progress has been made in actuator control and sensor integration.

Simulating robotic hands and integrating them into virtual environments is also becoming increasingly important. NVIDIA’s Isaac Sim platform, as detailed in their Robotics Fundamentals Learning Path, allows developers to import and simulate robots using the Unified Robot Description Format (URDF). This enables hands-on experience with configuring robotic assets and setting up control interfaces in a safe and controlled environment. Simulation is crucial for testing algorithms, optimizing designs, and reducing the cost and risk associated with physical prototyping.

The development of robotic hands isn’t limited to established companies. Individual makers and hobbyists are also contributing to the field. A Reddit post from April 4, 2021, showcased a robotic hand built using an Arduino, potentiometers for position sensing, and styrene sheet for the frame. While a prototype, it demonstrates the accessibility of robotics technology and the potential for innovation outside of traditional research institutions.

Despite the challenges, the pursuit of dexterous robotic hands continues. The ongoing research and development, coupled with advancements in materials, actuators, sensors, and control systems, suggest that more capable and versatile robotic hands will emerge in the coming years. These advancements will not only benefit the robotics industry but also have broader implications for automation, manufacturing, and human-robot interaction.

The current focus on development kits, like those offered by Shadow Robot, indicates a strategic approach: building a foundation of dexterity research before attempting large-scale deployment. This iterative process, combining academic research, commercial trials, and individual innovation, is likely to be key to unlocking the full potential of robotic hands.