Researchers develop low-cost sensor to enhance robots’ sense of touch

Researchers from Queen Mary University of London, in collaboration with teams from China and the USA, have developed an innovative L3 F-TOUCH sensor to enhance the tactile capabilities of robots. This sensor enables robots to “feel” objects and adjust their grip accordingly, a crucial step towards achieving human-level dexterity in manipulation and grasping.

The study, published in IEEE Robotics and Automation Letters, highlights the importance of reliable tactile information and force sensing in robot hands and graspers. The human hand possesses the ability to sense pressure, temperature, texture, and pain, as well as distinguish objects based on their shape, size, and weight. In contrast, many existing robot hands lack integrated haptic capabilities, making it challenging for them to handle objects effectively. Without a sense of touch and knowledge about the interaction forces, robots may struggle to maintain a secure grip on objects or handle delicate items without causing damage.

Led by Professor Kaspar Althoefer, the team presents the L3 F-TOUCH sensor, which stands for Lightweight, Low-cost, and wireLess communication. This high-resolution fingertip sensor offers a solution for measuring an object’s geometry and determining the forces required for interaction. Unlike other sensors that estimate interaction forces using camera images of soft elastomer deformation, the L3 F-TOUCH directly measures interaction forces through an integrated mechanical suspension structure with a mirror system, achieving higher accuracy and a wider measurement range.

The compact suspension structure allows the elastomer to deform upon contact with the surface, measuring high-resolution contact geometry exposed to external forces. The elastomer’s displacement is tracked using a special marker, enabling the measurement of contact forces along three major axes (x, y, and z) through a calibration process.

Professor Althoefer emphasizes that future work will focus on extending the sensor’s capabilities to measure rotational forces, such as twist, while remaining accurate and compact. This advancement will enable robots to have a more comprehensive sense of touch and improve their performance in manipulation tasks, including human-robot interaction scenarios like patient rehabilitation or assisting the elderly.

The breakthrough achieved with the L3 F-TOUCH sensor holds significant potential for the future of robotics. By providing robots with a sense of touch, they can handle objects more effectively and perform complex manipulation tasks with greater precision and reliability. This technology opens up new possibilities for more advanced and agile robots, ushering in a new era of robotic capabilities.

Posted in

Aihub Team

Leave a Comment





Reinforcement Learning: Training AI Agents to Make Decisions

Reinforcement Learning: Training AI Agents to Make Decisions

Natural Language Processing Unleashing the Power of Text

Natural Language Processing Unleashing the Power of Text

How AI is Transforming Industries

How AI is Transforming Industries

Exploring Neural Networks and Deep Learning

Exploring Neural Networks and Deep Learning

Ethical Considerations in Artificial Intelligence

Ethical Considerations in Artificial Intelligence

Computer Vision and Image Recognition in AI

Computer Vision and Image Recognition in AI

ARTIFICIAL INTELLIGENCE IN LOGISTICS

ARTIFICIAL INTELLIGENCE IN LOGISTICS

On Artificial Intelligence - A European approach to excellence and trust

On Artificial Intelligence – A European approach to excellence and trust

AI in Healthcare Advancements and Applications

AI in Healthcare Advancements and Applications

AI in Financial Services: Opportunities and Challenges

AI in Financial Services: Opportunities and Challenges

AI in Customer Service: Improving User Experience

AI in Customer Service: Improving User Experience

AI and Robotics: Synergies and Applications

AI and Robotics: Synergies and Applications

AI and Data Science: Bridging the Gap

AI and Data Science: Bridging the Gap

Top 10 emerging AI and ML uses in data centres

Top 10 emerging AI and ML uses in data centres

Piero Molino, Predibase: On low-code machine learning and LLMs

Piero Molino, Predibase: On low-code machine learning and LLMs

OpenAI’s first global office will be in London

OpenAI’s first global office will be in London

OpenAI is not currently training GPT-5

OpenAI is not currently training GPT-5

Microsoft’s AI chatbot is ‘unhinged’ and wants to be human

Microsoft’s AI chatbot is ‘unhinged’ and wants to be human

Machine learning expert Jordan bemoans use of AI as catch-all term

Machine learning expert Jordan bemoans use of AI as catch-all term

ITN to explore how AI can be a force for good at the AI & Big Data Expo this November

ITN to explore how AI can be a force for good at the AI & Big Data Expo this November

Fiverr create Demand for AI expertise surges by 1,000%

Fiverr create Demand for AI expertise surges by 1,000%

Databricks acquires LLM pioneer MosaicML for $1.3B

Databricks acquires LLM pioneer MosaicML for $1.3B

AI think tank calls GPT-4 a risk to public safety

AI think tank calls GPT-4 a risk to public safety

AI vs Machine Learning

AI vs Machine Learning

US: AI Begins Taking Over Thousands of Human Jobs | Vantage on Firstpost

US: AI Begins Taking Over Thousands of Human Jobs | Vantage on Firstpost

Snowpark, Input Tables, & Sigma AI: The Future of Analytics

Snowpark, Input Tables, & Sigma AI: The Future of Analytics

How to Scale Service with Generative AI and Einstein GPT

How to Scale Service with Generative AI and Einstein GPT

Fight AI with AI: Going Beyond ChatGPT

Fight AI with AI: Going Beyond ChatGPT

Can China’s ChatGPT clones give it an edge over the U.S. in an A.I. arms race?

Can China’s ChatGPT clones give it an edge over the U.S. in an A.I. arms race?

What Is AI Artificial Intelligence What is Artificial Intelligence

What Is AI Artificial Intelligence What is Artificial Intelligence