



Title: ALGORITHM IMPROVES SAMPLING OF CONTINUOUS DATA STREAMS
Inventors: Byung Lee, Mohammed Al-Kateb, and Sean Wang
Many sensor-based electronic devices collect samples of streaming data and store them in limited memory. Ideally, the sampled data accurately reflects the distribution of values in the entire stream.
Benefits of Adaptive Sampling
Often data storage capacity is shared among applications, such as a GPS device that can play MP3 files simultaneously. If it can adapt to changing demands for storage, when MP3 playback ends the memory available for GPS will increase, allowing more sampling of data and more accurate determination of geographic coordinates. In other devices, multiple data sources may compete for storage, such as hospital equipment that monitors several patients simultaneously. When monitoring is no longer needed for one patient, more data samples could be collected from the remaining patients.
Ensuring Sample Uniformity
Until now, it has not been possible to ensure uniform sampling of continuous data if the sample size adapts to available storage space. The algorithm presented here corrects this deficiency by ensuring uniformity when the sample size changes. If storage decreases, the algorithm randomly evicts data to maintain a uniform sample. If storage increases, the algorithm finds the minimum amount of data that should be drawn from the data stream to fill the storage, as well as the amount to randomly evict. In each case, uniformity is maintained because every possible value has an equal probability of being included in or rejected from the stored sample data with a required confidence level.
Commercialization
Industries that make use of sensor-based devices, such as heathcare, government, weather services, and transportation, would benefit from improved data sampling accuracy. Hardware manufacturers and software developers specializing in each of these fields could easily gain a competitive advantage and marketing opportunities by incorporating the algorithm into their products.
Competitive Advantages
- Optimizes memory usage in sensor-based devices.
- Increases accuracy of sampled data.
- Applicable to healthcare, government, transportation, and other industries.
- Adds value without increasing manufacturing costs
Patent/Licensing Status
Patent pending. Worldwide rights available.
Title: ADVANCED DATA MINING TOOL QUERIES MULTIPLE DATABASES
Inventor: Xingquan Zhu, Xindong Wu
In both business and research, valuable information can be found by applying complex queries to multiple databases. Typically this is done by querying each database separately and then analyzing the results for possible meta patterns.
The sequential approach is of limited value because the process is time consuming and often fails to recognize important patterns.
New Capabilities
The new approach proposed here easily performs complex multiple-database queries. It produces results many times faster than traditional methods and delivers more comprehensive results. Our system finds, analyzes, and develops rules for data patterns as the query proceeds. In effect, it learns as it goes to refine and streamline the search process. This reduces query time while revealing relational data that would otherwise be hidden. In one recent test, investigators ran queries on various types of U.S. Census data from multiple states. To generate 80,000 patterns, the traditional approach required 10,000 seconds, or more than 2.5 hours. In contrast, the integrated approach proposed here required only 106 seconds, or less than 2 minutes.
Commercialization
We anticipate strong demand in fields ranging from business and healthcare to higher education and government. Potential applications include retail sales, statistics, manufacturing, financial analysis, and sciences such as epidemiology and molecular biology. A commercial database developer could incorporate this technology into its product line at low cost and with minimal staff or programming time. The developer should realize a significant competitive advantage in each of its markets.
Next Steps
The technology has been tested successfully on databases of up to 250,000 records. Further testing is planned for databases of up to 1 million records.
Competitive Advantages
- Powerful tool for business intelligence and scientific analysis.
- Makes complex, multiple-database queries efficient.
- Reveals patterns that would otherwise remain hidden.
- Adds value for corporate, education, and healthcare markets.
- Low cost to implement in commercial databases.
Patent/Licensing Status
Patent pending. Worldwide rights available.
Title: ERS STANDARD CALCULATION FOR EM-MS
b
Inventors: Dennis Clougherty, Ph.D., et al.
Description: A major challenge posed in the assessment of risk at remediation sites and when detecting explosives is the ability to detect, quantify and determine the structure of compounds and their metabolites. Precise structural identification of chemicals is crucial, because chemical structure dictates biological, toxicological, and explosive activity. Moreover, accurate quantification is important to assess hazard and estimate risk. The coupling of electron monochromators to mass spectrometers (EM-MS) provides a potentially powerful tool for such molecular identification. EM-MS expands the sensitivity of traditional MS through the inclusion of a new dimension in the space of molecular characteristics - the electron resonance energy (ER) spectrum. However, before this technique can fulfill its full potential a library of ER spectra standards must be obtained in order to associate a spectrum with the molecule or class of molecules from which it arose, a likely cost-prohibitive venture. This invention provides a method to obtain these standards and enabling EM-MS by supplementing direct measurement with chemical inference and quantum scattering.
Advantages: Creation of a library of ER spectra for all molecules of potential interest is cost-prohibitive due to the large number of standards needed. The direct calculation method provided here offers a reduced cost method to obtain these spectra as needed and to quickly enable this important and powerful tool.
Patent Status: Patent pending
Licensing status: Worldwide rights available
Title: AKBRA - ADAPTIVE KNOWLEDGE-BASED RUN-TIME AGENTS
Inventors: Byung Lee, Firooz Sadjadi, and Sean Wang
Description: Coding of task specific run time code for remote devices has until now been mission specific and cumbersome to change. In addition, the small storage space on remote devices requires simple code, limiting the creation of more adaptive devices. AKBRA provides a multistage methodology for facilitating run-time code on a remote device quickly and flexibly when needed. In addition, it limits the constraints on the coding and on running the code in such devices. This methodology can be used in a variety of applications such as surveillance device deployment, including military and homeland security border control applications; tele-medicine for remote physician assistance and ambulance care; stock trading assistance; and traffic control.
Advantages: Current methods for managing remote devices have been task specific and unable to handle the variety and multitude of situations that would arise and need to be managed in a fully capable system. AKBRA provides the capability to expand remote systems from single to multi-task operations flexibly and quickly. In addition, the expediency and flexibility of this methodology reduces the costs of remote mission tasking.
Patent Status: Patent pending
Licensing status: Worldwide rights available
Title: SYSTEM AND METHOD FOR ANALYZING THE RESPONSE OF A MECHANICAL SYSTEM
Inventor: Jeffrey Frolik
Description: It is often desirable to analyze the response, or output, of a mechanical system to any one or more of a variety of inputs. Testing of a mechanical system is performed for a variety of reasons, including, determination of system performance, maintenance requirements, data collection for calibration of the system, or for simulation purposes. The invention is a system and method for analyzing the response of a mechanical system to a variety of one or more inputs. The use of several accelerometers have been operatively coupled in a way that collects the data required to perform these tasks and integrating this data into signal processing software. An added feature is that a wireless transmission of data can be configured for in situ monitoring.
Advantages: The advantage of this system is that the system response can be characterized/monitored in situ. For example, the performance of a shock absorber can be obtained using naturally occurring loads both by the environment and the rest of the mechanical system. This performance can be monitored over time to indicate when the shock should be replaced. Alternative approaches would replace shocks at scheduled intervals and not when necessarily needed.
Patent Status: U.S. patent application
Licensing status: Worldwide rights available
Title: ADAPTIVE PROTECTIVE COVERING FOR A CONDUCTOR
Inventors: Dryver Huston, Brian Esser
Description: This invention addresses the problem of abrasion, chafing and/or wear relative to various types of conductors. Currently, after installation of a conductor, vibration, movement and shifting of conductors through fittings, bulkheads and other structures cause harmful wear and abrasion, penetrating the exterior of the conductor and resulting in the failure of the system. The covering of the current invention serves to reduce the chafing, wear and abrasion by using an adaptive design that reacts to wear. This includes an outer moisture-proof flexible outer layer, with an additional layer of soft, flexible material, which further cures, hardens and toughens, in the presence of moisture .
Advantages: The requirement that conductors be routed through complex geometries dictates that the cables/tubing must be flexible during the installation process. This prohibits the use of very hard, durable coverings. This invention overcomes this limitation, by causing the cable to toughen in a fitting when subject to wear, reducing the chance of failure. However, because the covered cable only responds in the areas of wear, the rest of the cable remains flexible. This self-healing nature of the conductor can eliminate the need for bulkhead grommets and other conductor protecting devices. Applications for this technology include automotive, aerospace, construction and military markets.
Patent Status: Provisional patent
Licensing status: Worldwide rights available
watch the video on self-healing cable insulation
Title: DISCRETE-TIME ANALYTIC SIGNAL GENERATION SYSTEM
Inventor: Gagan Mirchandani, Mohammed Elfataoui
Description: Hilbert transformers have many applications in signal processing A common type of efficient communication scheme is Single-Side Band (SSB) Transmission. This is obtained from Double Side-Band (DSB) signals. Hardware, in the form of Hilbert transformers on a chip, exist for this (and many other) purposes. Since the bandwith of the SSB signal is ideally, half that of the DSB signal, good recovery of the original signal is possible. Also, again for the ideal case, one can safely sample by ½ (decimation). This means lower power consumption in the transmission process. In practice, the SSB signal is not fully single-side band since it does have non-zero spectra in the negative frequencies. Consequently, the exact signal may not be recovered and decimation will cause aliasing.
The invention, based on theoretical analysis and design, provides a way of designing Hilbert transformers that yield SSB signals that have greater attenuation of the spectrum in the negative frequency range. Accordingly, a better recovery of the signal and less aliasing after decimation becomes possible.
Advantages: The advantage of the new design is a ``more truer’’ SSB signal. Hence better reconstruction. Also when decimating, there is less aliasing and hence a better reconstruction of the original signal. In addition, decimation means smaller sized data and hence less memory and lower power requirements. The invention would apply to a number of important applications . Some of the key ones would be: DSP/SSB convertor, instantaneous frequency estimation, feature detection and signal denoising. The whole (single scale) concept can be extended to a multiscale or multiresolution context. See: http://www.emba.uvm.edu/~mirchand/temp/temp20.html
Patent Status: Patent pending
Licensing status: Worldwide rights available
Title: INTEGRATED MEMS-BASED ANTIVIBRATION DEVICE
Inventor: Dryver Huston, Brian Esser, James Plumpton
Description: The invention described here is a MEMS based vibration reduction device. The device utilizes both electrical and mechanical components, fabricated in a single package or as separate modules, which are connected together to form an active damping device. The integrated system in this document is referred to as the ‘AMD module.’ The AMD module can be attached to any component in which limiting the magnitude of vibration is desired. The item to which the device is attached in this document is referred to as the ‘base.’
There are several main components of this system, as described below:
1. Accelerometer: A MEMS accelerometer measures the magnitude and phase of vibration of the device and its base. These vibration measurements are used in the setup, tuning, operational and performance evaluation modes of operation of the AMD module.
2. Signal Processor/Microcontroller: The signal processor/microcontroller (SPM) is a component that takes information from the accelerometer and processes it as necessary to send a signal to the active mass actuator. The SPM can operate with either analog or digital techniques. A digital processing scheme utilizes an analog to digital converter to digitize the signals from the accelerometer. The SPM uses as inputs the amplitude and phase components of the measured acceleration signal, and possibly uses operational signals such as from an external controller, neighboring AMD modules, power supply, temperature, active mass actuator state and system tuning parameters. The SPM provides as output control signals to the active mass actuator, and possibly updated system tuning parameters, internal state values to external controllers and neighboring AMD modules.
3. Active Mass Actuator: This device can be made with any of the fabrication techniques described below, and will consist of a movable mass that can move within the AMD module in order to cancel out the inertial effects of the vibration by moving in a controlled manner out of phase relative to the motion of the base to reduce the amplitude of vibration.
The active mass could be free to move within a cavity, be attached with flexure mounts, or attached with other mechanical means. Actuation of the moving mass component can be performed with several methods. These include, but are not limited to, electromagnetic, electrostatic, pressure actuation, or thermal actuation. Control signals sent from the signal processor are used to control the active mass. Also, position based feedback may be used to indicate to the SPM control system precisely where the active mass component is at any time during operation. The active mass actuator can be either contained within a cavity, or may be exposed to the outside environment. Many geometric actuator configurations and actuator masses are possible, encompassing a wide range of applications. As the inertial effects are due to both the stroke and mass of the actuator, these two items can be tailored for specific applications and design constraints. Actuator mass can range from a few micrograms to kilograms for large applications. Similarly, strokes can range from micrometers for vary small high frequency applications to centimeters or even tens of centimeters for extremely large applications.
4. Power Supply: Power is required to operate the accelerometer, SPM controller, active mass actuator and any communications with external controllers and other AMD modules.
Power for the system can be supplied through cabling to other power supplies within the mounting infrastructure, may be from batteries mounted to the AMD module, or may also be through energy harvesting techniques in which the vibration itself is used to generate power and drive the mass component. Other powering techniques may include solar, chemical, mini fuel cells, or others.
Application: This device has many applications any place vibration may be a concern. This includes aerospace electronics, commercial electronics, satellite and space applications, precision metrology or production machinery, such as that use in semiconductor manufacturing industry. The device can be directly mounted to electronic circuit boards (ECB’s), where the power can be derived from the ECB’s power supply. It may also be attached in any vibration critical location on a wide variety of platforms, including disk drives, laser aiming devices, holographic storage writing mechanisms, gyroscopic devices, CD or DVD reader/writers, or others.
This invention not only can reduce the vibrations in new installations of the device, but can be used to reduce vibration isolation components already utilized to protect vibration sensitive applications. An example of this is the application to aerospace electronics, where integration of this device to existent or newly manufactured ECBs will reduce the stringent mounting requirements for the components, which will result in overall system weight savings with improved performance. Also, the device can be used to reduce vibrations in weight and space critical applications, such as the launch environments of space vehicles and satellites.
Other potential applications of the device are the reduction of trembling in human applications. For example, mounting the device inside binoculars or small arms can reduce the effect that human trembling has on the performance of these devices. Similarly, the devices could be mounted to cameras, which are excited by wind vibration, or to cameras mounted to to other devices which themselves are vibration sources (such as airplanes). Multiple devices can be arranged in an orthogonal configuration to reduce vibrations in multiple dimensions, as would be required for many applications.
The device can be mounted in a housing that is directly attached to a person’s appendage to reduce the appearance of trembling. This would be useful in combating the visual effects of diseases such as Parkinson’s syndrome, where sufferers often exhibit “shaky hands.” Having an un-obtrusive wrist mounted unit, inside a watch for example, may allow for much finer motor control of the hands by reducing the vibrations associated with muscle trembles.
Patent Status: US Patent Pending
Title: ON-OFF MAGNETIC ATTACHMENT DEVICE
Inventor: Dryver Huston, Brian Esser, James Plumpton
Description: This invention is intended to be used in applications where it is desirable to attach and remove a device without having to overcome the force of a permanent magnet or use a switchable magnet. Unlike electromagnets, devices according to the present invention are able to hold objects indefinitely without requiring power. Examples of specific applications include the use of a robotic foot, an attachment for sensors or transducers, or a magnetic pickup device. The invention utilizes a rotating cylindrical dipole magnet contained in a specially designed casing. The rotating magnet is caused to rotate by an acutator that is connected to the magnet. The magnet/casing housing becomes an effective gripper when it is placed near to or in contact with a ferromagnetic object. Rotating the cylindrical dipole magnet about the cylindrical axis through 90 degrees switches the magnet/casing between two states – on and off. Placing the magnet/casing on or off the surface of a ferromagnetic object switches the device between two more states. This combination gives four possible states of varying energy levels. The management of switching from state to state provides the sliding and gripping action.
Advantages: Switching the device between the four states in an automated fashion is a breakthrough that allows the rotation of dipole magnets and to move the magnet/casing closer to or away from the ferromagnetic object. Using multiple devices enables deployment of sensor systems that rely on the relative position of the sensors. This could serve as a robot foot for crawling up a ferromagnetic surface. Other applications include pickup devices, sensor deployment, and sensor attachment devices
Patent Status: Provisional Patent
Licensing status: Worldwide rights available
Title: ACTIVE VIBRATION DAMPING FOR CIRCUIT BOARDS
Inventor: Dryver Huston, Brian Esser, James Plumpton
Description: This invention is a method for actively damping the vibrations of electronic circuit boards (ECBs). ECBs that are used in spacecraft, aircraft and missile systems are exposed to severe shock and vibrations. Vibration frequencies seen in such applications can range from 3 to 5000 Hz, with acceleration levels from 1G to 30Gs (#1, pg11). The volume and weight of these electrical systems is crucial to their design. Electrical boxes, housing many ECBs are often strangely shaped to fit in small available spaces. The weight of any component used on spacecraft and aircraft are critical to fuel consumption and cost. Traditionally, ECBs that are exposed to such environments are designed to be very robust to prevent failure. Bulky mounting fixtures and reinforcing ribs are often used to reduce deflection and stresses in the boards due to vibration. Entire electronic subassemblies containing many ECBs may also be mounted with vibration isolators. The problem with these reinforced ECBs and vibration isolators for large boxes is that they consume too much space and add considerable weight to electrical systems. Active damping systems offer the opportunity to considerably reduce ECB vibrations by using active mechanical devices to damp the vibrations. There is a potential for a large reduction in weight relative to the use of passive systems.
The operating principle of an active mass damper is that it counteracts the vibrations of an object by driving a small mass in a motion that is out of phase with the vibrations of the object. The inertia forces generated by the active mass quell the vibrations. The main components of the system are shown in Figure 1. These are: 1. Active mass This is a mass that is attached to the vibrating structure. The mass of the active mass usually ranges from 1 to 10% that of the vibrating structure. 2. Actuator This is an electromechanical device that drives the active mass out of phase with the vibrations. The actuator will most likely be either a voice coil electromagnetic device, or a piezoelectric device. 3. Transducer This will sense the vibrations of the structure. It will most likely be a MEMs-based accelerometer. 4. Signal Processor The vibrations measured by the transducer will be processed to produce a control signal for the actuator and the inertial mass. The signal processor can be a simple operational amplifier or microprocessor. 5. Power amplifier The control signals from the signal processor will be amplified to power the actuator and to drive the inertial mass. The use of an active damper can result in overall structural weight reductions of 10 to 30%.
Advantages: This appears to be the first application of active vibration damping to the problem of excessive vibrations in circuit boards. A variety of passive techniques are already used to dampen excessive vibrations. The use of active damping offers the opportunity to reduce vibrations with devices that weigh less than the passive systems. Weight reduction can be very valuable in certain applications, such as aerospace electronics
Patent Status: Provisional Patent
Licensing status: Worldwide rights available
Title: AUTOMATED FRINGE COUNTING AND PROPERTY CALCULATION SYSTEM FOR BULGE TESTER
Inventor: Dryver Huston, Peter Sontag, Wolfgang Sauter
Description: This invention is a computer- based system that automatically controls the operation of a bulge tester, while counting interference fringes from a Michelson interferometer, and automatically calculating material properties from the fringe count. This invention allows manufacturers to automatically determine the material properties of the thin films such as curvature measurement and nanoindentation.
Applications: Bulge testing is a method of determining the material properties of thin films. Thin films are used in a variety of devices, including microelectronic circuits (computer chips), optical coatings, and protective wear resistant coatings such as bearing races.
Advantages: The bulge testing system is superior to existing methods such as curvature measurements and nanonindentation because of the convenience in operation and control.
Patent Status: Provisional Patent
Licensing status: Worldwide rights available
Title: STRUCTURE AND METHOD FOR ABRUPT PN JUNCTION DIODE FORMED USING CHEMICAL VAPOR DEPOSITION PROCESSING
Inventor: Walter Varhue
Description: The invention relates to forming a pn junction in a semiconductor structure, and, in particular, to forming a hyper-abrupt pn junction by low-temperature growth of a doped film on a surface of an oppositely doped substrate wherein the metallurgical junction of the pn junction could coincide with the surface of the substrate. The process makes it possible to place the metallurgical junction directly at the wafer surface and permits the growth of films of any doping concentrations, including doping concentrations in which the least doped side is greater than 1016 cm-3. The present process has been developed for use in an electron cyclotron resonance (ECR) PECVD reactor. However, the deposition conditions are also possible in other chemical vapor deposition apparatus.
Advantages: An advantage of the process of the present invention is its ability to produce properly functioning pn junction devices having the original wafer surface contained within the depletion region. This ability permits the diodes to be fabricated at any doping concentration, without concern for trap-assisted tunneling. This process can also be used to produce properly functioning devices having their metallurgical junctions coincident with the wafer surface. Epitaxial layers on a doped substrate can be grown at any doping concentration. The process may be performed at a lower temperatures to reduce or eliminate diffusion of dopant atoms across metallurgical junctions, while producing a hyper-abrupt pn junction.
Patent Status: US Patent applied for 6/00
Licensing status: Worldwide rights available
Title: PLASMA ENHANCED CVD PROCESS FOR RAPIDLY GROWING SEMICONDUCTOR FILMS
Inventor: Walter Varhue
Description: The invention relates to the epitaxial growth of semiconductor films and in particular to the rapid growth of silicon-based and germanium-based films. The deposition of appropriate materials on silicon or germanium wafers can be achieved at low substrate temperatures, i.e. below 650°C, and at high deposition rates, i.e., greater than 150 Å/minute, and as high as 500 Å/minute. The deposition process is assisted by a gaseous discharge, which acts to increase the deposition rate and to reduce the required substrate temperature. An exemplary gas discharge is an electron cyclotron resonance plasma, but this process in practice is not limited to such, and is extendable to other known plasma generating mechanisms, such as transformer coupled, inductively coupled, helicon, helical resonator, and remote or magnetically enhanced RF processes where energetic ion bombardment of the substrate surface is possible.
Advantages: This technology represents a breakthrough in the silicon production process by significantly reducing production temperature, while dramatically increasing growth rate. The process has been validated in Dr. Varhues laboratory.
Patent Status: US Patent applied for 9/14/99
Licensing status: Worldwide rights available
____________
UVM Innovations
University of Vermont, Office of Technology Transfer (OTT)
Todd.Keiller@UVM.edu
Available Technologies
New/Featured Technologies A/V Showcase Biomedical/Health Sciences
Agriculture/Life Sciences Engineering
About OTT
News FAQ Success Stories Staff/Contact Info
Working with OTT Corporate Registry
Licensing Services
How to License UVM Technologies Sample Forms and Agreements
_______________