Editorial Feature

Finders Keepers: RFusion Unveiled as New Household Search Robot

Researchers from the Massachusetts Institute of Technology (MIT) have unveiled a new domestic robot designed to help you find and collect lost items. The RFusion robot employs a radio frequency (RF) antenna and camera attached to a gripper at the end of its robotic arm.

Finders Keepers: RFusion Unveiled as New Household Search Robot

Image Credit: nd3000/Shutterstock.com

The RFusion system combines antenna signals with visual input from the camera to help locate and retrieve items, even if they are hidden out of view or buried beneath other objects.

In a paper delivered to the Association for Computing Machinery’s conference, Embedded Networked Sensor Systems, in 2021 and subsequently reproduced on MIT’s website, engineers demonstrated how they designed, implemented, and evaluated the RFusion system.

The system was designed to search for and retrieve items with RFID tags on them. It includes a robotic arm that carries the camera and antenna around a gripper at its endpoint.

It features two important innovations: geometrically fusing RF with visual information to improve target object location accuracy, even when the target object cannot be clearly seen, and a new reinforcement-learning artificial intelligence (AI) network that locates, goes toward and grasps target objects using RF-visual data efficiently.

To demonstrate the system, the engineers built an end-to-end prototype model of RFusion and tested it in challenging real-world conditions. Evaluation shows the localization technique developed for RFusion was capable of accuracy down to centimeter-scale and 96% success rate for retrieving objects, even when the objects were piled beneath other items.

The RFusion system could pave the way for more complex picking and retrieval tasks for robots in the future. Better picking robots could help with simple everyday domestic tasks like finding car keys, tidying up children’s toys, providing mobility and home care assistance, and so on. In industry, picking robots have applications in warehousing, logistics, manufacturing, and construction –any sector requiring automated mobility or manipulating objects.

Relying on RFID

The cutting-edge RFusion prototype gains a lot of usefulness from some relatively old-fashioned technology: RFID tags.

RFID tags are cheap, do not require any power source, scanning, and are readily available to manufacturers.

They work by reflecting signals that are emitted by special antennae. These RF signals, being of very low energy levels, can travel through a great deal more materials than optical radiation frequencies can.

This is what enables the RFusion system to detect and locate items even if they are occluded from view, finding a set of keys in a pile of clothing, for example.

Machine learning (ML) algorithms optimize the robot’s controls to automatically zero in on objects’ exact locations as efficiently as possible. Then, a grabber picks up the object and offers it up to the camera and antenna (all mounted on the same arm) for verification.

Without the ML algorithms to optimize the robot’s movements, it would have to swing the arm around the room to continually record measurements to determine the object’s exact location, a slow and inefficient process.

Using a type of ML called reinforcement learning, the researchers trained a neural network to optimize the RFusion robot’s route to the object. Reinforcement learning trains the network to achieve set goals through trial and error and a reward system. This works by allowing the AI to try different iterations, with mistakes punished and good actions rewarded.

To train RFusion, the AI was rewarded if it could reduce the number of moves the robot had to make to get to the item and the overall distance traveled.

All parts of the RFusion system – robotic arm, antenna, camera, grabber, and AI – are fully integrated. This means that it does not require any particular setup to work in any environment.

Broad Applications for RFusion System

The domestic applications for RFusion have captured popular interest, but the technology developed by MIT engineers may also have broad and impactful applications in industry in the future.

Logistics companies could use the system to sort through piles or layers of stock for order fulfillment in a warehouse. In manufacturing, it can be employed to identify and install components. There are also care applications for RFusion technology, both in patients’ homes and medical settings.

Fadel Abib, director of the Signal Kinetics group in the MIT Media Lab and associate professor in MIT’s Department of Electrical Engineering and Computer Science, was a senior author on the RFusion project. He said in an MIT press release:

“This idea of being able to find items in a chaotic world is an open problem that we’ve been working on for a few years. Having robots that are able to search for things under a pile is a growing need in industry today. Right now, you can think of this as a Roomba on steroids, but in the near term, this could have a lot of applications in manufacturing and warehouse environments.”

To make the RFusion system work in these potential near- and long-term applications, the system will have to be optimized for more speed. The current prototype, while well suited to help with daily tasks around the home, is not quite fast enough for reliable use in industrial applications like these.

Continue reading: Household Robotics: Enabling Innovation or Promoting Domestic Detachment?

References and Further Reading

Boroushaki, T. et al. (2021). RFusion: Robotic Grasping via RF-Visual Sensing and Learning. Proceedings of the 19th ACM Conference on Embedded Networked Sensor Systems. Available at: https://doi.org/10.1145/3485730.3485944

Zewe, A., (2021). A robot that finds lost items. Eurekalert.org. [online] Available at: https://www.eurekalert.org/news-releases/930589.

Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.

Ben Pilkington

Written by

Ben Pilkington

Ben Pilkington is a freelance writer who is interested in society and technology. He enjoys learning how the latest scientific developments can affect us and imagining what will be possible in the future. Since completing graduate studies at Oxford University in 2016, Ben has reported on developments in computer software, the UK technology industry, digital rights and privacy, industrial automation, IoT, AI, additive manufacturing, sustainability, and clean technology.

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Pilkington, Ben. (2021, December 30). Finders Keepers: RFusion Unveiled as New Household Search Robot. AZoRobotics. Retrieved on October 13, 2024 from https://www.azorobotics.com/Article.aspx?ArticleID=448.

  • MLA

    Pilkington, Ben. "Finders Keepers: RFusion Unveiled as New Household Search Robot". AZoRobotics. 13 October 2024. <https://www.azorobotics.com/Article.aspx?ArticleID=448>.

  • Chicago

    Pilkington, Ben. "Finders Keepers: RFusion Unveiled as New Household Search Robot". AZoRobotics. https://www.azorobotics.com/Article.aspx?ArticleID=448. (accessed October 13, 2024).

  • Harvard

    Pilkington, Ben. 2021. Finders Keepers: RFusion Unveiled as New Household Search Robot. AZoRobotics, viewed 13 October 2024, https://www.azorobotics.com/Article.aspx?ArticleID=448.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this article?

Leave your feedback
Your comment type
Submit

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.