Five patents that can help you save time,
energy—and even your cool
ibm-page-tools
Last year, IBM become the first company to surpass 4,000 patents in a single year
This marked the 16th consecutive year of U.S. patent leadership. While many of these patents will go into developing our solution portfolio, some are licensed to other companies for development and others are donated to commons.
Here is a look at five inventions inspired by the way we live every day. Traveling with babies in carseats. Taking pictures on vacation. Talking to a call center. And our researchers' ideas for making these experiences better.
Inventor: Laurence V. Marks
Suppose you are on vacation in a remote spot and see a beautiful sunset. The sun has just sunk below the horizon, the sky is purple and pink. In a moment or two, you could see the elusive "green flash." You reach for your camera and get the dreaded warning "Memory full." You don't have the time to go through and delete photos-most of them you want to keep anyway. What do you do?
There is always room for one more
This is the scenario IBM inventor Laurence Marks uses as the rationale for his new invention: adaptive compression quality. It takes advantage of the fact that in common compression techniques you can always compress an object to become smaller if you are willing to accept some loss of quality. The patent lays out a variety of ways in which the following can happen: When your camera's storage fills to a pre-set threshold, previously captured images are compressed and, in some cases, recompressed to a greater degree, to make room for new photos.
The sequence of compression steps is programmed onto a semiconductor chip or magnetic disk within the camera. It can occur automatically as a function of the storage, or the initial and maximum levels of compression/quality of image can be set by the user. What this patent advances, according to Marks, is the automatic switching of the levels of compression.
It can also be used for video and audio recording, for example in stretching the capacity of our voicemail so you never hear the words: "This mailbox is full." Now if we could just do the same thing for our own memories.
Inventors: Gregory Jensen Boss, Christopher James Dawson, Peter George Finn, Rick Allen Hamilton II
Today, a visually impaired or blind person generally has a choice between using a seeing-eye dog or a cane to help navigate his environment. This patent, based on RFID technology, can provide a viable third alternative.
It describes a system in which the person is able to receive the information he needs, via an electronic, synthesized voice, to safely make his way through a room, a store or classroom or down a street.
The system relies on electronic RFID tags situated throughout the environment that transmit information to a battery-powered tag reader, which the user can carry or wear. Each tag can provide extensive detail about an object. Such technology is used today in inventory control, automated tool booths and security systems. However, if it was applied for use by a visually impaired person, he would be overwhelmed by data since there is no means of prioritizing the information.
What this patent does is set up a framework in which the user creates rules about how much information he receives from each of the tagged items and in what order so that he can regulate the inflow of data within an environment. For example, the system can prioritize objects into three levels of safety, differentiating an electrified fence at the highest level while designating a telephone post at the second level. To alert a person to the fence, a verbal and tactile alert may be programmed into the system, while for the telephone post there may simply be a verbal warning.
Within the system, there can be an unlimited number of levels and gradations within levels as well as different categories of information, such as path hazards, food expiration dates, pharmacy locations. Almost as important as which information to include is which data to suppress.
Inventors: Paul A. Moskowitz, Rick A. Hamilton II, Brian M. O'Connell, Clifford A. Pickover, Keith R. Walker
For many households, the first step toward "going green" is changing the light bulbs from incandescent light bulbs to compact fluorescent light bulbs (CFLs). This is because CFLs use 75 percent less energy and last about ten times longer.1 But what do you do if you're in charge of changing light bulbs for an entire company? Or a town?
This is what IBM inventor Paul Moskowitz wondered about as he watched the staff at the IBM Yorktown Research building change the light bulbs in the dining hall over time. He noticed that some had not been changed because the view of the bulbs themselves was obscured by lamp shades.
Similarly, many bulbs in street lamps and office buildings are encased in a frosted envelope or located in light fixtures so far away that it is hard to distinguish by either color or quality of the light what sort of bulbs they are. Moskowitz and four colleagues have patented a methodology for a device that provides for the "remote inspection of light sources to determine the type of bulbs." It outlines two approaches.
Measure the flickering
Incandescent lights emit a relatively constant light by means of a heated filament, whereas fluorescent bulbs emit a flickering light. Microprocessor logic can be programmed to detect variations in light intensity (flicker) based on actual measurements.
Separate it into a spectrum
The other method is to use an apparatus that separates the light into its spectral components to determine whether there are well-defined spectral lines (CFL bulbs) or the spectrum is continuous (incandescent bulbs).
The detection component can be built into a small, handheld portable device outfitted with wireless capabilities and a laser pointer for ease of use. To scale the system, the output of the microprocessor can be combined with a GPS or satellite wireless system to detect and record, for example, the location of street lamps that need to be changed. Or the data can be streamed to a carbon offset provider, which could analyze light ratios and determine ways to offset energy use, trigger a warning, or issue an incentive for the organization to take action.
1 http://www.energystar.gov/index.cfm?c=cfls.pr_cfls
Inventors: Joseph Kinman Lee, Daniel Ray Spach, Prasanna Srinivasan, Darren Paul Umstead
A parent forgets his child in the back seat of a car and the child dies of hyperthermia. This devastating accident happens 15 to 25 times a year2 in the United States. It is also a relatively recent phenomenon tracing to the early 1990s when car-safety experts began to recommend that car seats be moved to the back seat, and that baby seats be reversed to face the rear of the car-away from the driver, reducing the overall visibility of the child.
"My first child was about to be born when I heard about such a case," remembers IBM engineer Daniel Spach, inspiring him to devise an invention that uses RFID and wireless technology to detect-and ultimately protect-a baby in a car seat. The idea is to embed RFID tags into the car seat at the manufacturing site. When a consumer buys a car seat, the embedded tag can then be registered to a car equipped with an RFID reader and control unit.
The RFID tag can detect the presence of the child via the seat buckle or by registering her weight in the seat. The RFID reader is mounted in the car cabin along with a main control unit that is programmed to read and activate RFID signals according to a set of timed protocols.
A sample scenario would be that after a set lapse of time, if the child is still detected in the seat, an alarm would sound. Further measures could include rolling down the window after a set period of time; adjusting the air conditioning if the car is running and even dialing 911 if it has a communications system.
2 "Fatal Distraction," The Washington Post, 03/08//09
Inventors: Sara Basson, Dimitri Kanevsky, Nitendra Rajput, Alexander Faisman
The client speaks English. The two experts speak Russian and Chinese. The subject is a technical issue related to an optical instrument. Time is of the essence. And it all takes place over the phone.
As businesses operate more globally, they will increasingly handle complex issues among multilingual clients, colleagues and partners. A team of IBM inventors has developed a methodology that knits together a variety of technologies including speech recognition, language/accent identification, translation, transcription and database analytics to create a more productive experience in call centers, for example.
We speak your language
"The basic idea behind the patent is how do we do a better job of making sure that you are connected with someone who, if possible, is from your language group, and to do it in a way that is transparent," explains Sara Basson, director of IBM's speech transcription strategy. "The first step is automatic language or accent identification, so that you get routed directly to a person who is available and speaks your dialect."
The patent goes on to outline a number of scenarios. "While call centers are clearly an area, these technologies are also relevant to governmental organizations and universities," says Basson. "In fact, this particular patent evolved from discussions with a colleague in India, about issues they were having in providing parliamentary speeches to citizens." At current count, India has 22 official regional languages3 and 122 spoken by 10,000 people or fewer.4
Call centers have tiered levels of assistance. Once the caller has been matched to a tier-one operator who understands the question, it may be determined she needs to be escalated to a tier-two operator, who happens to speak only Chinese. At this point, a machine translation component could kick in.
The accent is on accuracy
Given that the nature of voice recognition and translation is not perfect, a restating technique might be used by a human expert or an automated system to help improve accuracy of the text before it gets escalated. This means that the statement is reworded in a simpler form, changing the passive voice to active, using smaller sentences and replacing confusing words.
Dimitri Kanevsky is a master inventor who develops new algorithms for improving accuracy in speech recognition and translation and has received more than 100 patents. For truly complex issues, he says, multiple experts in different languages could be assigned to work on different parts of the same question. "For example, one question can be answered by someone in China, another part of the question by Russia," he explains. "The translation engine integrates all of the messages so that everyone can understand what is being is talked about and the question won't have to be repeated again and again."
3 http://languages.iloveindia.com/
4 http://en.wikipedia.org/wiki/Languages_of_India
IBM Research goes into making our planet smarter
Here's how
