As people increasingly use robots in various applications requiring precise accuracy, some are interested in technologies that could improve how those advanced machines work. There are many examples of how targeted, high-tech applications could help you or others developing new systems.

1. Digital Twins

Developing robots in the lab is an excellent first step, but how can those working on such projects verify that their creations have the necessary control mechanisms to navigate diverse real-world scenarios?

A Singapore University of Technology and Design team deployed a digital twin within robot simulation software to determine how the machines would handle different built environments. Figuring out those details in advance is crucial since many people use robotic solutions in settings ranging from industrial sites to office buildings.

One prong of the group’s approach concerned assessing how well existing built environment guidelines cater to robots’ needs. Next, those involved built digital twins reflecting robot archetypes and example environments to see whether challenges arose.

Additionally, since the digital twin simulations supported hazard navigation, people could identify potential obstacles. Knowing about those challenges before sending the robots into the real world allowed developers to alter their programming decisions. The associated outcomes could result in better robotic control system accuracy.

The digital twin simulations let researchers examine path planning, navigational capabilities and environmental interactions. Such evaluations are also valuable for showing people which types of robots work best for where they plan to use them. In one case, the group tested four cleaning robots in six settings. The results showed which machine achieved the most goals and demonstrated the best overall performance.

Those working on this project believe it will shape future design guidelines and improve the introduction of robots into urban environments. Since many people plan to use these machines in crowded environments or those with uneven terrain, those developing them can save time by learning as much as possible about likely challenges in advance.

2. Advanced Sensors

People also explore how innovative sensors could make robotic control systems more accurate. One case comes from the Korea Advanced Institute of Science & Technology. Researchers focused on sensors associated with wearable robots. Ongoing investigations indicate they could be instrumental in specific health care cases, such as stroke rehabilitation. This group wanted to design possibilities with control capabilities that stayed reliable despite contact with the wearer’s sweat and dead skin.

They created a stretchable electromyographic sensor that can detect physiological signals without showing degraded performance due to moisture on the skin’s surface. This project’s participants confirmed that existing sensors often showed less accuracy over time and had decreased performance due to skin deformation. However, the researchers dealt with those challenges by designing their sensors with microneedle arrays that stayed in place without causing discomfort.

Lab results suggest these sensors would allow the dependable control of wearable robots, enabling people to keep using them to get high-quality electrophysiological signals from users. Such feedback is vital for these machines to interpret someone’s movement intentionality correctly.

Another example of the need for robotic technology with highly precise sensors comes from an exoskeleton that helps people with lower-limb disabilities to walk. Statistics showed that people demonstrated a 51% increase in gait ability after three months of using this device.

Parties going through health care rehabilitation programs appreciate seamless robotic interactions. Additionally, their providers will be more likely to trust the technology if they believe its control systems will work well under demanding real-world conditions. Thoughtfully designed sensors can achieve the desired performance and create additional use cases.

3. Reinforcement Learning

Reinforcement learning is an artificial intelligence training technique that teaches algorithms to respond in specific desirable ways to get rewards. AI systems trained this way can become highly adaptive, making them well-suited to industrial environments and other busy settings.

Some decision-makers achieve better precision in other ways. For example, robots controlled with digital servo motors can display positional accuracy down to fractions of degrees, while control algorithms maintain performance despite load variations and other disturbances.

However, researchers wondered whether a reinforcement learning-based approach applied to servo motors could improve operating consistency by further elevating control system precision. Staying ahead of the curve often means being open to continuous improvement opportunities, as this team was. It is also important that people tailor their research to emerging areas, such as Industry 4.0. Then, those efforts can gain traction as the public recognizes their relevance.

This group evaluated their innovation on three servo motors, observing how it impacted various tasks. The results showed it offered superior torque and speed control performance compared to its counterparts. Tests confirmed how the servo’s algorithms could respond to target trajectory changes, promptly adjusting the motor’s running state.

Although the participants found that the algorithm optimized the servo motor in dynamic conditions, they clarified that future studies should include methods to reduce the scale of their neural networks. Such progress would make their algorithms more computationally efficient.

4. Virtual Reality

Virtual reality has gained significant momentum, becoming widely adopted to advance workforce training, mental health treatments and much more. Could it positively affect how people control robots, too? Those taking part in a collaboration between Massachusetts Institute of Technology and the University of California at San Diego hoped to find out.

Their work centered on robotic teleoperation, allowing people to control the machines without being physically near them. Some companies offer these solutions so people can continue attending work or school if injuries, chronic illnesses or other challenges prevent them from going in person. A common option is to attach a screen to a remotely controlled robot on wheels. Then, colleagues, classmates and others can see who owns and moves the machine.

This joint effort between two universities introduced their control method called Open-TeleVision. Those who developed this option say it creates a much more immersive experience for the person piloting the robot.

The individual begins by strapping on a VR headset, which allows them to see through the robot’s eyes with a stereoscopic perspective. Then, they move their bodies to reflect how they want the robotic machine to move.

Experiments showed people’s experiences mimicked how they would feel if they were actually performing tasks such as picking up balls and putting them into cups, even though the robot did all the work. Researchers envision various potential applications for their control method, ranging from search and rescue operations to planetary exploration.

Technologies Shape Robotic Control Systems

These fascinating examples show some of the advancements at the heart of modern robotic control systems. Besides enabling the case studies described here, what people learn through these developments will undoubtedly inform related progress.

By: DocMemory
Copyright © 2023 CST, Inc. All Rights Reserved