Projects

(ARCHIVED AND NO LONGER UPDATED)

SoberDiary (Ongoing) :: behavior sensing & feedback, alcohol addiction 

SoberDiary is a phone-based support system that helps alcohol-dependent patients to maintain sobriety in their diarly lives once they have completed alcohol withdrawl treatment. By carrying a Bluetooth breathalyzer that is wirelessly connected with the SoberDiary application installed on their phones, patients can perform breath alcohol tests to self-monitor their own alcohol use. To motivate behavioral change, SoberDiary also records personal progressand achievements in maintaining sobriety for patients review. SoberDiary also incorporates the principles of the 12-step program to enable patients to learn the concepts associated with each step, thereby promoting self-fulfillment to recognize the values and beliefs of the community and the self. All test results and momentary feedback sampled by the SoberDiary client are used to provide appropriate recommendations for reducing alcohol cravings, and is uploaded to a backend server to enable continuous patient monitoring.

Bioscope (Ongoing) :: sensing, health

BioScope is an an extensible bandage system with components that can be stacked like Lego blocks. Using this system, healthcare workers can simultaneously collect the four most commonly monitored biosignals (i.e., heart rate, body temperature, acoustic signals emitted from the body, and inertial readings of human movement) from multiple bandages to assess and diagnose physical conditions. BioScope extracts the processing and communication functions into a core building block, and hosts the required sensors. Each sensor is affixed as a patch that collects one biosignal. By stacking the required sensors onto a bandage-like platform, healthcare workers can easilycreate a customized bandage that can be affixed to the skin of the patient. The data collected by the sensors are sent through a Bluetooth interface to the device screen used by the healthcare worker.

SensingFork (2014) :: behavior sensing & feedback, children

The Sensing Fork system contains a digital fork and a mobile interactive and persuasive game for a young child who is a picky eater and/or easily distracted during mealtime. To sense a child's eating behavior, we have designed and prototyped a sensor-embedded digital fork. To detect the food color in contact with the fork, a small color sensor is embedded at the tip of the Sensing Fork device. The Sensing Fork has other sensory and communication components including three electrodes, a motion sensor, a microcomputer, a Bluetooth radio, and a battery. By processing the sensory data, the fork can recognize a child's eating behavior (eating actions and food colors). The sensory data is transmitted to the Hungry Panda game executed on a smartphone. The game is interactive and provides feedback to encourage children to improve their eating behavior.

SmartTooth (2013) :: sensing, health

Smart tooth is an in-mouth oral sensory system, which uses a small accelerometer sensor embedded inside artificial teeth. Our motivation was based on our observation that most oral activities, such as chewing, drinking, speaking, and coughing, each produce a unique teeth motion. By recording and identifying teeth motion profiles for each oral activity, the proposed oral sensory system builds classifiers that distinguish different human oral activities.

HeatProbe (2011) :: sensing, energy

HeatProbe is a thermal-based power meter system that tracks the energy consumed by individual appliances. HeatProbe disaggregates total power usage from a master pow-er meter into individual appliance power usage by using a thermal camera and image processing techniques to detect the heat patterns generated by running appliances. Since appliances produce heat as a byproduct of their operation, HeatProbe determines changes in an appliance's power state by tracking its surface temperature. A rising surface temperature implies that the appliance has been turned on, while a falling surface temperature implies the appliance has been turned off.

PipeProbe (2010) :: sensing

PipeProbe is a mobile sensor system that can determine the spatial topology of hidden water pipelines behind walls. PipeProbe works by dropping a tiny wireless sensor capsule into the source of the water pipelines. As the PipeProbe capsule traverses the pipelines, it gathers and transmits pressure and angular velocity readings. Through temporal-spatial analysis on the sensor readings, our algorithm locates all turning points in the pipelines and maps their 3D spatial topology.

Double-side Multi-touch (2009) :: interaction technique

This is a new mobile interaction model, called double-side multi-touch, based on a mobile device that receives simultaneous multi-touch input from both the front and the back of the device. This new double-sided multi-touch mobile interaction model enables intuitive finger gestures for manipulating 3D objects and user interfaces on a 2D screen.

Playful Bottle (2009) :: behavior sensing & feedback, health

The Playful Bottle explores the use of a mobile phone, when attached to an everyday object used by an everyday behavior, becomes a tool to sense and influence that behavior. This mobile persuasion system, called Playful Bottle system, makes use of a mobile phone attached to an everyday drinking mug and motivates office workers to drink healthy quantities of water. A camera and accelerometer sensors in the phone are used to build a vision/motion-based water intake tracker to detect the amount and regularity of water consumed by the user. Additionally, the phone includes hydration games in which natural drinking actions are used as game input. Two hydration games are developed: a single-user TreeGame with automated computer reminders and a multi-user ForestGame with computer-mediated social reminders from members of the group playing the game.

SpinLoc (2008) :: sensing, localization

This work proposes the novel use of spinning beacons for precise indoor localization. The proposed "SpinLoc" (Spinning Indoor Localization) system uses spinning (i.e., rotating) beacons to create and detect predictable and highly distinguishable Doppler signals for sub-meter localization accuracy. The system analyzes Doppler frequency shifts of signals from spinning beacons, which are then used to find the direction from the spinning center to the target. By obtaining direction of the target from two or more spinning beacons, SpinLoc can precisely locate stationary targets. After designing and implementing the system using MICA2 motes, its performance was tested in an indoor garage environment. The experimental results revealed a median error of 40~50 centimeters and a 90% error of 70~90 centimeters.

ChroMirror (2008) :: sensing

This study presents the ChroMirror system, a digital mirror imaging system which helps users select appropriate clothing color combinations. By digitally rendering a mirror image of the user wearing clothing in different color-harmonic combinations, this system enables users to easily and playfully explore a wide variety of chromatic and color-harmonic clothing combinations (i.e., without having to physically change clothing). A long-term goal of the ChroMirror system is to encourage users to experiment with colorful and color-harmonic clothes.

Nutrition-aware kitchen (2007) :: behavior sensing & feedback, health

This work is a smart kitchen with UbiComp technology to improve home cooking by providing calorie awareness of food ingredients used in prepared meals during the cooking process.The kitchen has sensors to track the number of calories in food ingredients, and then provides real-time feedback to users on these values through an awareness display.

Mug Tree (2007) :: behavior sensing & feedback, health

We have explored the design of a playful mug, called Mug-Tree, to motivate people to drink water regularly and to develop a good water-drinking habit. Our system includes (1) a smart mug that can recognize user drinking from it, and (2) a digital photo frame that displays a playful game connecting water drinking to watering a virtual tree. The Mug-Tree reminds users to drink water regularly, and also help users to develop a good water drinking habit by attracting continuous participation to this game.

Playful Toothbrush (2007) :: behavior sensing & feedback, health

It is about a playful toothbrush to assist parents in motivating and getting their young children into a habit of proper and thor-ough tooth brushing. Our system includes a vision-based motion tracker that recognizes different tooth brushing motions, and a fun tooth brushing game in which a young child clean his/her mirror dirty virtual teeth by physically brushing his/her own teeth.

Adaptive RIP System (2007) :: sensing, localization

It is about enhancing the accuracy of radio interferometric positioning (RIP) system by adaptively selecting the best set of beacon nodes for locating a set of mobile targets. To realize this adaptive RIP system, we have developed an estimation error model to predict positioning error of the RIP algorithm given different combinations of beacon nodes. Building upon this estimation error model, we further devise an adaptive algorithm that finds the optimal beacons that give a smallest amount of positional error to a set of mobile targets.

Persuasive Lunch Tray (2006) :: behavior sensing & feedback, health

It is an interactive, persuasive game built into an ordinary lunch tray to assist parents to improve dietary behaviors of their young children. The persuasive game is played over a smart lunch tray. By eating from the lunch tray, a child can see his/her favorite cartoon character being colored. The smart lunch tray incorporates both the context-awareness (of pervasive computing) and the persuasive media (of persuasive computing), enabling the creation of a smart object that is not only aware of human behavior but can also influence and shape human behaviors through their natural interactions with the object.

Energy-efficient Zigbee Localization (2006) :: sensor network, localization

It is an energy-aware indoor localization system based on Zigbee radio sensor network. Given a request for a certain positional accuracy from an application, it can meet this request while minimizing energy consumption on a target mobile badge. Our method is to adapt the sampling rate to the target's mobility level. Our localization method is based on combining signal strength fingerprinting and signal strength propagation model, optimizing it for Zigbee radio characteristics. We have created several real testbed deployments, and we have shown that energy saving can be as high as 50%.

Object Location Tracker (2006) :: sensing, localization

Do you have any frustrating experiences in which you forget where you last placed things, such as glasses, cell phone, wallet, keys, remote controls, etc.? This is an object locator that can automatically track whereabouts of objects you have misplaced by hands, so you won't waste your time looking for misplaced objects. This system is composed of (1) a finger ring RFID antenna, (2) a wrist band connecting the RFID reader to a Zigbee radio sensor node, (3) RFID tags on tracked objects, and (4) any choice of an indoor localization system (see the energy-efficient Zigbee localization above).

Diet-aware Dining Table (2005) :: sensing, health

It is a dietary tracker built into an ordinary dining table. It is a "smart" dining table that is "aware" of our natural eating behavior. It can automatically track what and how much we eat from the tabletop surface. The goal is to provide users with information about their eating patterns, therefore, help them sensible and healthy eating. The dining table is augmented with two layers of weighting and RFID sensor surfaces to detect and recognize multiple, concurrent person-object interactions occurring on the table.

Geta Sandals (2005) :: sensing, localization

It is a self location tracker built into an ordinary Japanese Geta sandals, making them "smart" sandals aware of their user's location. It is based on a footprint-based indoor location system. It works by measuring and tracking the displacement vectors along a trial of footprints (each displacement vector is formed by drawing a line between each pair of footprints). The position of a user can be calculated by summing up the current and all previous displacement vectors. In comparison to existing indoor location systems, this footprint location system has a unique advantage that it is infrastructure-free. A user simply has to wear the Geta sandals to track his/her locations without any setup or calibration efforts. This makes our footprint method easy for everywhere deployment.

Sensor-assisted Wi-Fi Indoor Location System (2005)

Wi-Fi based indoor location systems have been shown to be both cost-effective and accurate, since they can attain meter-level positioning accuracy by using existing Wi-Fi infrastructure in the environment. However, two major technical challenges persist for current Wi-Fi based location systems: instability in positioning accuracy due to changing environment dynamics, and the need for manual offline calibration during site survey. We have looked at three dynamic environment factors (people, doors, and humidity) that can interfere with radio signals and cause positioning inaccuracy in the Wi-Fi location systems are identified. We have proposed a sensor-assisted adaptation method that employs RFID sensors and environment sensors to adapt the location systems automatically to the changing environment dynamics. The adaptation method performs online calibration to build multiple context-aware radio maps under various environment conditions, selects the radio map that best matches the current environment condition, and uses it for to determine the location.

mProducer (2004) :: personal experience computing

The mProducer is a mobile authoring tool that enables everyday users to effectively and efficiently perform archiving and editing at or immediately after the point-of-capture of digital personal experiences from their camera-equipped mobile devices.