National Science Foundation (NSF)

Subterms

    More stories

    • 75 Shares109 Views

      in Security

      A robot that finds lost items

      by

      A busy commuter is ready to walk out the door, only to realize they’ve misplaced their keys and must search through piles of stuff to find them. Rapidly sifting through clutter, they wish they could figure out which pile was hiding the keys.

      Researchers at MIT have created a robotic system that can do just that. The system, RFusion, is a robotic arm with a camera and radio frequency (RF) antenna attached to its gripper. It fuses signals from the antenna with visual input from the camera to locate and retrieve an item, even if the item is buried under a pile and completely out of view.

      The RFusion prototype the researchers developed relies on RFID tags, which are cheap, battery-less tags that can be stuck to an item and reflect signals sent by an antenna. Because RF signals can travel through most surfaces (like the mound of dirty laundry that may be obscuring the keys), RFusion is able to locate a tagged item within a pile.

      Using machine learning, the robotic arm automatically zeroes-in on the object’s exact location, moves the items on top of it, grasps the object, and verifies that it picked up the right thing. The camera, antenna, robotic arm, and AI are fully integrated, so RFusion can work in any environment without requiring a special set up.

      While finding lost keys is helpful, RFusion could have many broader applications in the future, like sorting through piles to fulfill orders in a warehouse, identifying and installing components in an auto manufacturing plant, or helping an elderly individual perform daily tasks in the home, though the current prototype isn’t quite fast enough yet for these uses.

      “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,” said senior author Fadel Adib, associate professor in the Department of Electrical Engineering and Computer Science and director of the Signal Kinetics group in the MIT Media Lab.

      Co-authors include research assistant Tara Boroushaki, the lead author; electrical engineering and computer science graduate student Isaac Perper; research associate Mergen Nachin; and Alberto Rodriguez, the Class of 1957 Associate Professor in the Department of Mechanical Engineering. The research will be presented at the Association for Computing Machinery Conference on Embedded Networked Senor Systems next month.

      Play video

      Sending signals

      RFusion begins searching for an object using its antenna, which bounces signals off the RFID tag (like sunlight being reflected off a mirror) to identify a spherical area in which the tag is located. It combines that sphere with the camera input, which narrows down the object’s location. For instance, the item can’t be located on an area of a table that is empty.

      But once the robot has a general idea of where the item is, it would need to swing its arm widely around the room taking additional measurements to come up with the exact location, which is slow and inefficient.

      The researchers used reinforcement learning to train a neural network that can optimize the robot’s trajectory to the object. In reinforcement learning, the algorithm is trained through trial and error with a reward system.

      “This is also how our brain learns. We get rewarded from our teachers, from our parents, from a computer game, etc. The same thing happens in reinforcement learning. We let the agent make mistakes or do something right and then we punish or reward the network. This is how the network learns something that is really hard for it to model,” Boroushaki explains.

      In the case of RFusion, the optimization algorithm was rewarded when it limited the number of moves it had to make to localize the item and the distance it had to travel to pick it up.

      Once the system identifies the exact right spot, the neural network uses combined RF and visual information to predict how the robotic arm should grasp the object, including the angle of the hand and the width of the gripper, and whether it must remove other items first. It also scans the item’s tag one last time to make sure it picked up the right object.

      Cutting through clutter

      The researchers tested RFusion in several different environments. They buried a keychain in a box full of clutter and hid a remote control under a pile of items on a couch.

      But if they fed all the camera data and RF measurements to the reinforcement learning algorithm, it would have overwhelmed the system. So, drawing on the method a GPS uses to consolidate data from satellites, they summarized the RF measurements and limited the visual data to the area right in front of the robot.

      Their approach worked well — RFusion had a 96 percent success rate when retrieving objects that were fully hidden under a pile.

      “Sometimes, if you only rely on RF measurements, there is going to be an outlier, and if you rely only on vision, there is sometimes going to be a mistake from the camera. But if you combine them, they are going to correct each other. That is what made the system so robust,” Boroushaki says.

      In the future, the researchers hope to increase the speed of the system so it can move smoothly, rather than stopping periodically to take measurements. This would enable RFusion to be deployed in a fast-paced manufacturing or warehouse setting.

      Beyond its potential industrial uses, a system like this could even be incorporated into future smart homes to assist people with any number of household tasks, Boroushaki says.

      “Every year, billions of RFID tags are used to identify objects in today’s complex supply chains, including clothing and lots of other consumer goods. The RFusion approach points the way to autonomous robots that can dig through a pile of mixed items and sort them out using the data stored in the RFID tags, much more efficiently than having to inspect each item individually, especially when the items look similar to a computer vision system,” says Matthew S. Reynolds, CoMotion Presidential Innovation Fellow and associate professor of electrical and computer engineering at the University of Washington, who was not involved in the research. “The RFusion approach is a great step forward for robotics operating in complex supply chains where identifying and ‘picking’ the right item quickly and accurately is the key to getting orders fulfilled on time and keeping demanding customers happy.”

      The research is sponsored by the National Science Foundation, a Sloan Research Fellowship, NTT DATA, Toppan, Toppan Forms, and the Abdul Latif Jameel Water and Food Systems Lab. More

    • 150 Shares169 Views

      in Environment

      Zeroing in on the origins of Earth’s “single most important evolutionary innovation”

      by

      Some time in Earth’s early history, the planet took a turn toward habitability when a group of enterprising microbes known as cyanobacteria evolved oxygenic photosynthesis — the ability to turn light and water into energy, releasing oxygen in the process.

      This evolutionary moment made it possible for oxygen to eventually accumulate in the atmosphere and oceans, setting off a domino effect of diversification and shaping the uniquely habitable planet we know today.  

      Now, MIT scientists have a precise estimate for when cyanobacteria, and oxygenic photosynthesis, first originated. Their results appear today in the Proceedings of the Royal Society B.

      They developed a new gene-analyzing technique that shows that all the species of cyanobacteria living today can be traced back to a common ancestor that evolved around 2.9 billion years ago. They also found that the ancestors of cyanobacteria branched off from other bacteria around 3.4 billion years ago, with oxygenic photosynthesis likely evolving during the intervening half-billion years, during the Archean Eon.

      Interestingly, this estimate places the appearance of oxygenic photosynthesis at least 400 million years before the Great Oxidation Event, a period in which the Earth’s atmosphere and oceans first experienced a rise in oxygen. This suggests that cyanobacteria may have evolved the ability to produce oxygen early on, but that it took a while for this oxygen to really take hold in the environment.

      “In evolution, things always start small,” says lead author Greg Fournier, associate professor of geobiology in MIT’s Department of Earth, Atmospheric and Planetary Sciences. “Even though there’s evidence for early oxygenic photosynthesis — which is the single most important and really amazing evolutionary innovation on Earth — it still took hundreds of millions of years for it to take off.”

      Fournier’s MIT co-authors include Kelsey Moore, Luiz Thiberio Rangel, Jack Payette, Lily Momper, and Tanja Bosak.

      Slow fuse, or wildfire?

      Estimates for the origin of oxygenic photosynthesis vary widely, along with the methods to trace its evolution.

      For instance, scientists can use geochemical tools to look for traces of oxidized elements in ancient rocks. These methods have found hints that oxygen was present as early as 3.5 billion years ago — a sign that oxygenic photosynthesis may have been the source, although other sources are also possible.

      Researchers have also used molecular clock dating, which uses the genetic sequences of microbes today to trace back changes in genes through evolutionary history. Based on these sequences, researchers then use models to estimate the rate at which genetic changes occur, to trace when groups of organisms first evolved. But molecular clock dating is limited by the quality of ancient fossils, and the chosen rate model, which can produce different age estimates, depending on the rate that is assumed.

      Fournier says different age estimates can imply conflicting evolutionary narratives. For instance, some analyses suggest oxygenic photosynthesis evolved very early on and progressed “like a slow fuse,” while others indicate it appeared much later and then “took off like wildfire” to trigger the Great Oxidation Event and the accumulation of oxygen in the biosphere.

      “In order for us to understand the history of habitability on Earth, it’s important for us to distinguish between these hypotheses,” he says.

      Horizontal genes

      To precisely date the origin of cyanobacteria and oxygenic photosynthesis, Fournier and his colleagues paired molecular clock dating with horizontal gene transfer — an independent method that doesn’t rely entirely on fossils or rate assumptions.

      Normally, an organism inherits a gene “vertically,” when it is passed down from the organism’s parent. In rare instances, a gene can also jump from one species to another, distantly related species. For instance, one cell may eat another, and in the process incorporate some new genes into its genome.

      When such a horizontal gene transfer history is found, it’s clear that the group of organisms that acquired the gene is evolutionarily younger than the group from which the gene originated. Fournier reasoned that such instances could be used to determine the relative ages between certain bacterial groups. The ages for these groups could then be compared with the ages that various molecular clock models predict. The model that comes closest would likely be the most accurate, and could then be used to precisely estimate the age of other bacterial species — specifically, cyanobacteria.

      Following this reasoning, the team looked for instances of horizontal gene transfer across the genomes of thousands of bacterial species, including cyanobacteria. They also used new cultures of modern cyanobacteria taken by Bosak and Moore, to more precisely use fossil cyanobacteria as calibrations. In the end, they identified 34 clear instances of horizontal gene transfer. They then found that one out of six molecular clock models consistently matched the relative ages identified in the team’s horizontal gene transfer analysis.

      Fournier ran this model to estimate the age of the “crown” group of cyanobacteria, which encompasses all the species living today and known to exhibit oxygenic photosynthesis. They found that, during the Archean eon, the crown group originated around 2.9 billion years ago, while cyanobacteria as a whole branched off from other bacteria around 3.4 billion years ago. This strongly suggests that oxygenic photosynthesis was already happening 500 million years before the Great Oxidation Event (GOE), and that cyanobacteria were producing oxygen for quite a long time before it accumulated in the atmosphere.

      The analysis also revealed that, shortly before the GOE, around 2.4 billion years ago, cyanobacteria experienced a burst of diversification. This implies that a rapid expansion of cyanobacteria may have tipped the Earth into the GOE and launched oxygen into the atmosphere.

      Fournier plans to apply horizontal gene transfer beyond cyanobacteria to pin down the origins of other elusive species.

      “This work shows that molecular clocks incorporating horizontal gene transfers (HGTs) promise to reliably provide the ages of groups across the entire tree of life, even for ancient microbes that have left no fossil record … something that was previously impossible,” Fournier says. 

      This research was supported, in part, by the Simons Foundation and the National Science Foundation. More

    • 138 Shares169 Views

      in Environment

      Making catalytic surfaces more active to help decarbonize fuels and chemicals

      by

      Electrochemical reactions that are accelerated using catalysts lie at the heart of many processes for making and using fuels, chemicals, and materials — including storing electricity from renewable energy sources in chemical bonds, an important capability for decarbonizing transportation fuels. Now, research at MIT could open the door to ways of making certain catalysts more active, and thus enhancing the efficiency of such processes.

      A new production process yielded catalysts that increased the efficiency of the chemical reactions by fivefold, potentially enabling useful new processes in biochemistry, organic chemistry, environmental chemistry, and electrochemistry. The findings are described today in the journal Nature Catalysis, in a paper by Yang Shao-Horn, an MIT professor of mechanical engineering and of materials science and engineering, and a member of the Research Lab of Electronics (RLE); Tao Wang, a postdoc in RLE; Yirui Zhang, a graduate student in the Department of Mechanical Engineering; and five others.

      The process involves adding a layer of what’s called an ionic liquid in between a gold or platinum catalyst and a chemical feedstock. Catalysts produced with this method could potentially enable much more efficient conversion of hydrogen fuel to power devices such as fuel cells, or more efficient conversion of carbon dioxide into fuels.

      “There is an urgent need to decarbonize how we power transportation beyond light-duty vehicles, how we make fuels, and how we make materials and chemicals,” says Shao-Horn, emphasizing the pressing call to reduce carbon emissions highlighted in the latest IPCC report on climate change. This new approach to enhancing catalytic activity could provide an important step in that direction, she says.

      Using hydrogen in electrochemical devices such as fuel cells is one promising approach to decarbonizing fields such as aviation and heavy-duty vehicles, and the new process may help to make such uses practical. At present, the oxygen reduction reaction that powers such fuel cells is limited by its inefficiency. Previous attempts to improve that efficiency have focused on choosing different catalyst materials or modifying their surface compositions and structure.

      In this research, however, instead of modifying the solid surfaces, the team added a thin layer in between the catalyst and the electrolyte, the active material that participates in the chemical reaction. The ionic liquid layer, they found, regulates the activity of protons that help to increase the rate of the chemical reactions taking place on the interface.

      Because there is a great variety of such ionic liquids to choose from, it’s possible to “tune” proton activity and the reaction rates to match the energetics needed for processes involving proton transfer, which can be used to make fuels and chemicals through reactions with oxygen.

      “The proton activity and the barrier for proton transfer is governed by the ionic liquid layer, and so there’s a great tuneability in terms of catalytic activity for reactions involving proton and electron transfer,” Shao-Horn says. And the effect is produced by a vanishingly thin layer of the liquid, just a few nanometers thick, above which is a much thicker layer of the liquid that is to undergo the reaction.

      “I think this concept is novel and important,” says Wang, the paper’s first author, “because people know the proton activity is important in many electrochemistry reactions, but it’s very challenging to study.” That’s because in a water environment, there are so many interactions between neighboring water molecules involved that it’s very difficult to separate out which reactions are taking place. By using an ionic liquid, whose ions can each only form a single bond with the intermediate material, it became possible to study the reactions in detail, using infrared spectroscopy.

      As a result, Wang says, “Our finding highlights the critical role that interfacial electrolytes, in particular the intermolecular hydrogen bonding, can play in enhancing the activity of the electro-catalytic process. It also provides fundamental insights into proton transfer mechanisms at a quantum mechanical level, which can push the frontiers of knowing how protons and electrons interact at catalytic interfaces.”

      “The work is also exciting because it gives people a design principle for how they can tune the catalysts,” says Zhang. “We need some species right at a ‘sweet spot’ — not too active or too inert — to enhance the reaction rate.”

      With some of these techniques, says Reshma Rao, a recent doctoral graduate from MIT and now a postdoc at Imperial College, London, who is also a co-author of the paper, “we see up to a five-times increase in activity. I think the most exciting part of this research is the way it opens up a whole new dimension in the way we think about catalysis.” The field had hit “a kind of roadblock,” she says, in finding ways to design better materials. By focusing on the liquid layer rather than the surface of the material, “that’s kind of a whole different way of looking at this problem, and opens up a whole new dimension, a whole new axis along which we can change things and optimize some of these reaction rates.”

      The team also included Botao Huang, Bin Cai, and Livia Giordano in the MIT’s Research Laboratory of Electronics, and Shi-Gang Sun at Xiamen University in China. The work was supported by the Toyota Research Institute, and used the National Science Foundation’s Extreme Science and Engineering Environment. More