mercredi 7 juillet 2004

'Agrobots' Go to the Farm Autonomous robots, dubbed 'ag robots,' are being developed at the University of Illinois at Urbana-Champaign (UIUC) by agricultural engineers to help farmers to do their duties. Some of them, 'ag ants,' costing only $150, walk through crop rows on their mechanical legs. When one of them detects weed plants, he alerts the other members of the robotic strike force to attack the plants as a team. These robots, and others developed at UIUC, have sensors to detect the end of crop rows, so they can automatically turn. The engineers also built a high-tech robot, which costs $7,000, uses a laser to estimate the distance to corn plants. Future versions will be used to detect plants diseases or to apply precise amounts of pesticides. Here is a photograph of an 'ag robot' (Credit: Tony Grift, UIUC). Here is the introduction of the news release. Farm equipment in the future might very well resemble the robot R2D2 of Star Wars fame. But instead of careening through a galaxy far, far away, these ag robots might be wobbling down a corn row, scouting for insects, blasting weeds and taking soil tests. University of Illinois agricultural engineers have developed several ag robots, one of which actually resembles R2D2, except that it's square instead of round. The robots are completely autonomous, directing themselves down corn rows, turning at the end and then moving down the next row, said Tony Grift, University of Illinois agricultural engineer. Grift said he's using a "smaller and smarter" approach to farming. The long-term goal, he said, is for these small, inexpensive robots to take on some of the duties now performed by large, expensive farm equipment. As Grift asked, "Who needs 500 horsepower to go through the field when you might as well put a few robots out there that communicate with each other like an army of ants, working the entire field and collecting data?" And speaking of ants, one of the robots coming out of ag engineering is a foot-long "Ag Ant", which is being designed to walk through crop rows on mechanical legs. Built for only $150, these cheap robots could someday be used to form a robotic strike force. "We're thinking about building 10 or more of these robots and making an ecosystem out of them," Grift said. "If you look at bees, they will go out and find nectar somewhere. Then a bee will go back and share this with the group and the whole group will collect the food. Similarly, one robot might find weed plants. Then it would communicate this location to the other robots and they would attack the plants together as a group -- an ecosystem, if you will." They also developed other kinds od farming robots. In addition to the Ag Ant, Grift and Yoshi Nagasaka, a visiting scholar from Japan, developed a more expensive, high-tech robot for about $7,000. This robot guides itself down crop rows using a laser mounted in front to gauge the distance to corn plants. Meanwhile, Grift and Matthias Kasten, an intern from Germany, have built yet another robot, this one for roughly $500. The robot is equipped with two ultrasonic sensors that bounce sound waves off of objects, as well as four of the cheap infrared sensors used in simple motion detection sensors. Now, what's next? Grift would like to someday see an experimental farm where all of the work is being performed by autonomous robots. And he said the logical place for such an ambitious farm would be Illinois. But right now, they're simply focusing on navigation skills for the robots. Eventually, these robots could be equipped to perform duties, such as detecting disease, weeds or insects, sampling soil or even applying pesticides. For more information about the subject, you can read a good introduction by Tony Grift, "Advanced Machinery for Biosystems Applications" (PDF format, 20 pages, 618 KB). This paper contains an history of how robots can be used in outdoor environments and several pictures of early agrobots. Sources: UIUC new release, July 6, 2004; and other UIUC pages 2:50:54 PM   Permalink   Comments []   Trackback []   Technorati about this page and this post

mardi 6 juillet 2004

Rent a Bicycle with your Cell Phone If you live in West London, more precisely in the Hammersmith and Fulham borough, you can use your mobile phone to unlock and rent a bicycle. In "Phones power bike rental scheme," BBC News Online writes that users receive a PIN via SMS when they want to rent a bicycle and a second one to return it. They also receive the price of the trip via SMS and are charged monthly on their credit cards. Rates are fairly low, from 30p for less than 15 minutes, up to £8 for one to eight hours. The company behind this new scheme, OYBike, has already 28 docking stations near metro stations or car parks, and expects to extend this service to other areas of London starting this month. Here are two small pictures of the special lock and bicycles (Credit: OYBike). Here is how the system works according to BBC News Online. Anyone wanting to use an Oybike signs up and pays a £10 annual fee, then uses their mobile phone to unlock a bike for as long as they need. The phone receives a pin number to unlock the bike at the start of a journey and a second one to lock it at the end and to determine how much the rider is charged. Hire prices start at 30p for 15 minutes and rise to £8 for up to eight hours of time on an Oybike. And your clothes will not get dirty when using these bicycles. Suit wearers will be happy to know that the Oybikes do not use chains, so there's little chance of getting oil on your clothes. OYBike claims that you'll save time and money by using this system. Here is an example. Bob lives in Queensway and works in Brook Green. He would normally get the Central line tube to Shepherds Bush, and then either get a bus or walk through Shepherds Bush Green, and along the Shepherds Bush Road for the remainder of his journey. OYBike could make his journey easier. Instead of having to wait for a bus, or take the time to walk he could rent an OYBike from outside of Shepherds Bush Central line tube station and cycle the 5 minutes to the office, leaving the OYBike at an empty docking station just outside the office in Brook Green. Here is the comparison betwween the bus and the bicycle rides. Cost of Journey via Bus: £0.70 Cost of journey via OYBike: £0.30 (as journey time is between 0 and 15 minutes) Estimated time of journey via OYBike: 5 minutes Estimated time of journey via bus: 20 minutes (including waiting time) This method of using a phone to reserve a bicycle was patented in August 2002. For more information, you can use the Online European Patent Register search engine. You just have to enter the application number "02250767.7" without quotes. For a quicker access, here is a direct link to the abstract of the patent, "Controlled access systems." If you are very curious, just hit the "Drawing" button. It's very refreshing to see that such a simple sketch can be used to deposit a patent in Europe. Sources: Mark Ward, BBC News Online, July 5, 2004; OYBike website; Online European Patent Register website 2:07:00 PM   Permalink   Comments []   Trackback []   Technorati about this page and this post

lundi 5 juillet 2004

True or False? Investigating Digital Images In our world of digital information, everything is described by zeros and ones, even photographs. And it's incredibly easy to use computers to alter an image after it is digitized to produce fake ones, like the recent forged image showing Jane Fonda and John Kerry together at a political meeting. Now, computer scientists from Dartmouth College have developed an algorithm able to tell the difference between a "real" image and a modified one. They "built a statistical model that captures the mathematical regularities inherent in natural images. Because these statistics fundamentally change when images are altered, the model can be used to detect digital tampering." The team thinks that their technology, or a similar one, will soon be incorporated in the U.S. legal system to authenticate images. Here is a description of the problem. A digital image is a collection of pixels or dots, and each pixel contains numbers that correspond to a color or brightness value. When marrying two images to make one convincing composite, you have to alter pixels. They have to be stretched, shaded, twisted, and otherwise changed. The end result is, more often than not, a realistic, believable image. "With today's technology, it's not easy to look at an image these days and decide if it's real or not," says Farid, an Associate Professor of Computer Science. "We look, however, at the underlying code of the image for clues of tampering." And here is brief description of the solution devised by Farid and graduate student Alin Popescu. Farid's algorithm looks for the evidence inevitably left behind after image tinkering. Statistical clues lurk in all digital images, and the ones that have been tampered with contain altered statistics. "Natural digital photographs aren't random," he says. "In the same way that placing a monkey in front of a typewriter is unlikely to produce a play by Shakespeare, a random set of pixels thrown on a page is unlikely to yield a natural image. It means that there are underlying statistics and regularities in naturally occurring images." Here are two links to the abstract and the full text (PDF format, 15 pages, 4.58 MB) of "Statistical Tools for Digital Forensics," a paper presented at the 6th International Workshop on Information Hiding, held in Toronto, Canada, in 2004. Here is the text of the abstract. A digitally altered photograph, often leaving no visual clues of having been tampered with, can be indistinguishable from an authentic photograph. As a result, photographs no longer hold the unique stature as a definitive recording of events. We describe several statistical techniques for detecting traces of digital tampering in the absence of digital watermarks or signatures. In particular, we quantify statistical correlations that result from specific forms of digital tampering, and devise detection schemes to reveal these correlations. Below is an example of a forgery. The images are extracted from the paper mentioned above (Credit: Hany Farid and Alin Popescu, Dartmouth College). Original image The forgery consists of removing a stool and splicing in a new floor taken from another image (not shown here) of the same room. And here is the estimated probability map (p) of the forgery. And here is Farid's conclusion. "This technology to manipulate and change digital media is developing at an incredible rate," says Farid. "But our ability to contend with its ramifications is still in the Dark Ages. I'm always asked if this technology would stand up in a court of law." He explains that the simple answer is, "eventually." Farid predicts there will be skepticism and a great deal of scientific and legal debate. But eventually, he believes that some form of his technology or someone else's will be incorporated into our legal system. If you are interested or intrigued by this subject, here are two other links to the abstract and the full text (PDF format, 11 pages, 6.50 MB), of another paper, "Exposing Digital Forgeries by Detecting Traces of Re-sampling" (currently available in the IEEE Transactions on Signal Processing). Sources: Dartmouth College Press Release, July 1, 2004; and various pages at Dartmouth College 1:19:50 PM   Permalink   Comments []   Trackback []   Technorati about this page and this post

dimanche 4 juillet 2004

Use DidTheyReadIt in France, Go To Jail For 5 Years! You might have heard about the Did They Read It? service, which promises to let you know if your email correspondents have opened, read or forwarded your messages (check "Tracking the e-mail you sent" for more details). If you live in France, I strongly urge you not to use this service. In this privacy alert (in French), the Commission Nationale de l'Informatique et des Libertés (CNIL) says that the service is illegal and breaches French privacy legislation. A French user of this service faces now up to five years in jail and a fine of 300,000 euros (about US$ 360,000). I would never have used this tracking service, but five years in jail seem ridiculous and excessive. What do you think? The European Digital Rights group, which has 14 members organisations from 11 different countries in Europe, commented this French decision in its last EDRi-gram bulletin. This is the item #6 of this bulletin. Here is the short EDRI-gram warning. The French data protection authority CNIL has declared the new U.S. mail-service 'Did they read it?' illegal. Through this service, launched in May 2004 by Rampell Software, subscribers get a report about the exact time their e-mail was opened, for how long, on what kind of operating system and if the mail was forwarded to other people. To use this service, subscribers simply forward their mail to Rampell, after which a one-pixel gif is added that allows for this kind of tracking. Rampell carefully avoids explaining the technology, and just promises that e-mails are being kept confidential. The CNIL finds the service unacceptable under the French privacy legislation of 1978. The recipients do not have a choice to accept or refuse sending this information to the sender and aren't even informed. Because the service provides detailed information on the reading behaviour, the data are considered sensitive, and the collection illegal. Any French subscriber to this service risks a prison sentence of 5 years and a fine of 300.000 euro. You'll find additional comments in this article from InternetWeek.com. "Did They Read It?" is certainly a questionable service -- and it doesn't even work for users who don't read their emails in HTML format -- but five years in jail for using it is outrageous and even ridiculous. Please let me know what you think. Sources: Mark Glassman, The New York Times, via International Tribune, June 5, 2004; CNIL, June 22, 2004; EDRI-gram, Number 2.13, June 30, 2004; InternetWeek.com, July 2, 2004 1:12:36 PM   Permalink   Comments []   Trackback []   Technorati about this page and this post

samedi 3 juillet 2004

A Video Projector that Fits in Your Pocket Video projectors able to project high-quality images will be embedded in your cellphones and laptops within two years. This is the promise of a new technology developed at Cambridge University. These pocket projectors will have no lenses and no light bulbs. Instead, these future battery-powered tiny projectors will rely on holographic technology and special algorithms. In "Holograms enable pocket projectors," Technology Research News explains that a 2D hologram will be created on a microdisplay and projected by using a laser beam. This has been possible because the researchers have written special algorithms which generates the holograms a million times faster than standard ones. Here is a description of the concept. Key to the device's diminutive size is the lack of lenses and high-power light bulbs. Conventional digital video projectors form images by generating a small picture on a transparent microdisplay inside the projector, then shining a high-power light through the microdisplay to a large magnifying lens. In the researchers' design, a two-dimensional hologram is shown on the microdisplay rather than an image, and the projected image is formed by shining a laser beam through the microdisplay, which scatters the light into a particular pattern. "No lenses are required -- the projected image is formed entirely by diffraction," said Adrian Cable, a researcher at the University of Cambridge in England. Another enabling technology is the use of special algorithms for a fast generation of the holograms. The researchers' algorithm solves the problem, said Cable. The key was studying how noise, or the distortion of an image, affects people's perception of video. The researchers showed test subjects 300 pairs of video images and asked them to rate which, if either, was higher quality. The images were the same but contained different levels of noise. The researchers found that variation in noise levels affected people's perception of video quality more than the actual level of the noise. The researchers' algorithm generates holograms about one million times faster than the standard direct binary search algorithm running on a 2 gigahertz Athlon personal computer, according to Cable. "The figure below shows a horse-faced cheat being attacked by a crazed half-man-half-beast (left) and the same image (right), but reconstructed from a computer-generated hologram. The hologram took a few seconds to generate using the OSPR algorithm on a 2 GHz PC, representing an orders of magnitude improvement in both execution time and image quality over conventional hologram generation algorithms." (Credit for images and legend: Edward Buckley, Cable's professor at the Cambridge University Engineering Department (CUED)). When will be be able to use our cellphones as projectors? The researchers' prototype is black and white. They are working on configuring two or more of the devices in parallel to generate full-color video. The researchers aim to produce practical pocket-sized video projectors in two to five years, said Cable. Sources: Eric Smalley, Technology Research News, June 30/July 7, 2004; Edward Buckley, Cambridge University Engineering Department 3:09:25 PM   Permalink   Comments []   Trackback []   Technorati about this page and this post

vendredi 2 juillet 2004

This New Radar Can See Through Walls According to Haaretz, an Israeli start-up has developed a new radar technology to see through walls. This radar system, based on UWB (ultra wideband) technology, can produce 3D images of what stays behind walls. The real breakthrough is that this system can be used from a distance of up to 20 meters, which will benefit rescuers as well as military personnel by providing useful information about the number of people inside a room, their locations and even their weapons. The newspaper adds that the images are of good quality, allowing the users of the system to follow what is happening behind the wall in real time. However, don't expect to get one today. The first devices are expected to be available within 18 months. Here are some excerpts. a small, Herzliya-based company called Camero is offering a solution: a radar system, based on UWB (ultra wideband) technology, that can produce three-dimensional pictures of what lies behind a wall, from a distance of up to 20 meters. The pictures, which resemble those produced by ultrasound, are relatively high-resolution. Although the figures are somewhat blurred, the system enables the user to follow what is happening behind the wall in real time. "The company was born of urgent operational needs," said CEO Aharon Aharon -- and not only those of the military. "When disaster victims must be rescued from a collapsed building or a fire, time is of the essence," he explained. "Rescue forces often invest enormous resources and precious time in combing the rubble, or endanger their lives by entering the flames, even if it is not clear that there are any survivors behind the walls." And here are some details about the technology, which also uses special software. Camero was born at the Jerusalem Global venture capital fund (JVG), when Amir Be'eri, a former defense establishment employee associated with the fund (his most recent position was CEO of Infineon), developed a way to emit UWB radio waves. UWB was a new technology at the time, and it was necessary because ordinary radio waves do not provide high enough resolution to be useful. Yet radio waves are necessary because other types of waves do not pass through walls. Another problem with radio waves is that they do not function well around metal. However, Camero has developed sophisticated software that enables its technology to work even on steel-reinforced concrete walls. Apparently, the first devices will be ready within 18 months, a period during which Camero's competitor, Time Domain, will be able to sell its own technology. Time Domain, which also uses UWB technology to see through walls, has been active for six months and is already selling millions of dollars worth of devices a year. But Camero's technology is superior in several important respects. First, it can be used from a distance of 20 meters, whereas Time Domain's product must be right next to the wall in question. Second, it gives a detailed picture of everything in the room, whereas Time Domain's product locates objects, but gives no information about their shape or size. In "Israeli invention sees through walls," WorldNetDaily gives some details about other defense technologies developed in Israel. Israeli firms are well known for developing revolutionary technology, particularly in the defense fields. El Al Airlines recently implemented a high-tech antimissile system developed by an Israeli firm, and Israel announced it developed a Star Wars-like remote control border with Gaza that uses unmanned sensor patrol cars and computerized observation posts to automatically spot and, upon human authorization, kill terrorists, even recommending the most appropriate weapon for the system to fire against a specified target. In addition, an Israeli security source told WND that Israel recently developed proprietary technology that can discreetly put an electronic field around a building or area that gives users the ability to monitor and control every electronic emission within that field, from electronic can openers to fax machines, computers and cell phones. Sources: Zuri Dar and Oded Hermoni, Haaretz, July 1, 2004; Aaron Klein, WorldNetDaily.com, July 2, 2004 1:58:16 PM   Permalink   Comments []   Trackback []   Technorati about this page and this post

jeudi 1 juillet 2004

'Facetop' Blends Screen and Video Images I'm sure that many of you have had poor experiences when participating to phone or video conferences. Now, a new video conferencing interface, named Facetop, improves the level of collaboration by blending transparent images of the user filmed by a video camera on the computer display. This results in a 'ghost' image of the user on the screen. When he points at something, "his video reflection appears to touch objects on the screen." The computer scientists also developed a two-user version in which the 'ghost' images of the two users appear side by side. Both can alternatively take control of the desktop, again allowing a better collaboration. You can expect a Mac version within months and a Windows version in two years. Researchers from the University of North Carolina at Chapel Hill have devised a videoconferencing system that comes a step closer. Facetop superimposes transparent images of a computer's desktop over video images of the user to allow the user to look at the video and desktop at the same time. The video shows a ghostly mirror image of the user so that when he points, his video reflection appears to touch objects on the screen. The system tracks fingertip position in the video to allow the user to control the mouse pointer. As it turns out, the human visual/brain system "seems to be quite good at paying attention to one and ignoring the other, depending on whether you want to see the user or the desktop information," said David Stotts, an associate professor of computer science at the University of North Carolina at Chapel Hill. Here you can see the 'ghostly' reflection of the user blended with what he sees on his screen. The two-user version combines the video streams of remote users and presents the images side-by-side. This allows each person to control the desktop, watch the other person control the desktop, and see the other's face. These two images are extracted from a technical report named "Support for Distributed Pair Programming in the Transparent Video Facetop" (PDF format, 10 pages, 683 KB). This report is written in almost plain english and contains lots of pictures. What kind of applications can we expect and is it an expensive one? The system can be used for remote teaching, PowerPoint presentations, and as a basic PC interface with fingertip and pointing mouse control, said Stotts. The system is relatively inexpensive to implement, said Stotts. "All we need is a $100 FireWire camera, a Macintosh and the Internet," he said. The system's software controls the transparency effect and the fingertip tracking. So when will we see such a system and what are the next steps? A Macintosh version of FaceTop could be implemented within a few months, and a PC version within two years, according to Stotts. The PC version would have to wait for the next major version of the Windows operating system, which will have the needed technical infrastructure, he said. The researchers are working on a multi-person version of the system, are performing studies of how well the system works for collaborative programming, and are adapting a version that can be used with wide, very high quality displays. Sources: Kimberly Patch, Technology Research News, June 30/July 7, 2004; University of North Carolina at Chapel Hill website 3:15:31 PM   Permalink   Comments []   Trackback []   Technorati about this page and this post

© Copyright 2004 Roland Piquepaille. Last update: 07/07/2004; 14:54:38.