Project Ideas

Here we have a list of ideas for projects that we would like to see worked on. These are suggestions only and we also had a large number of people attend class at the beginning of the semester suggesting ideas that are not on this list. These ideas are all meant as suggestions to help you figure out a direction for your team.

If you're contributing ideas to this list, make sure your idea has a title, contact info, and a short blurb describing it. It may also be a good idea to include some idea of what might be reasonably expected for a project, as well as what work has already been completed.

If you're looking to pick up one of these projects, talk to the person listed as the contact.

Photo Sharing for Police

Interview Atlanta police officers (or firemen) and determine how photo sharing during patrol/missions would be beneficial for them. Extend Mobiphos, a current application for the social sharing of photos on mobile phones written by last year's TA Nirmal Patel, to address the officer's needs. Initial interviews suggest adding a map and GPS information on where each photo was taken would be beneficial. (Jung Soo Kim, Thad Starner)

Telesign (Communication between the deaf and hearing)

Help perform a quantitative study on the current practices of how the deaf communicate with the hearing (i.e. pen & paper). Compare current practice to typing on a Blackberry and a sign→English phrase translator we have created. Help design the testing software and testing procedures. (Kim Weaver & Thad Starner)

Social Mobile Dance Games

Adapting Nirmal Patel's Mobile Dance Revolution game (DDR on a mobile phone), create new games where teams of players have to dance in synch to succeed. Use our Bluetooth accelerometers to sense new dance steps. Or create the mobile phone version of Rockband! Support new platforms, like the new OpenMoko phone or the iPhone

Rehabilitation games

Using wireless sensors (Wimotes, our Bluetooth accelerometers, wireless microphones, etc.) create games that promote rehabilitation. An ongoing project, AstroWheelie, embeds sensors in a wheelchair so that the user plays Asteroids by spinning in place (to spin the ship) and rocking back and forth (to fire). There are many opportunities. Improve and evaluate Astrowheelie. Use a microphone or strain gauges so that the player plucks the wheelchair's spokes as an interface. Create games that use walkers as controllers. Or create games for the elderly to keep fit (“whack a mole” with the feet, or peddling in play to play a racing game).


Recently, someone in Atlanta died because a deaf person could not make a phone call to 911 (emergency call centers can not receive SMS nor trace their location with e911). Last year, students created Deaf911 on a Nokia phone which emulates TTY on the voice lines so that a deaf user could contact the 911 center using the current technology. However, the Deaf use Blackberries and Sidekicks for their good keyboards. Port Deaf911 (now TTYPhone) so that the deaf can actually use it.

Subtle Gesture Interfaces

Can you control your computer with a wink of an eye or a flick of the wrist? Read Vernor Vinge's “Rainbow's End” for inspiration and then use our gesture toolkits to see if you can create an interface that is so subtle your friends don't think you are using it

Bluetooth keyboards

Help Steve Akins (a previous student) create the infrastructure to design new wireless keyboards and mice for Linux/Mac/Windows.

Piano learning

Suppose you want to learn the piano. Put on the Pianoglove and a mp3 player. Throughout your day (while reading email, working out, etc.) the system plays the song you want to learn and vibrates your fingers at the appropriate time for each note. Help Kevin Huang improve his hardware and do the testing to show that this tactile “passive learning” is possible for musical instruments.

3d scanner

Make a 3d scanner that can be used with our 3d printer. One of the challenges will be getting the scanned data into a format that is useful for editing and printing.

3d circuits

Experiment with creating 3d circuits using a sign cutter, the laser cutter, and/or the 3d printer and conductive inks.

Make a wearable

Improve our current gumstix or single board computer designs and make an inexpensive wearable. (Thad Starner, Stephen Cuzzort, Kent Lyons, Dan Ashbrook)

wii controller for video editing

Using a Wii controller, make a gesture system for editing video (Profesor Irfan Essa is interested in this one). See also the MythTV project suggested by a student below.

Speech enabled mobile web 2.0

Web sites are being designed for allowing users to annotate the site. Yet, when mobile, the user does not have the manual dexterity to type quickly. Make a web 2.0 toolkit for integrating mobile speech. The system should allow a mobile user to add a “voicenote” to the web page. The voicenote is kept in an audio file as well as a rough text translation that is indexable in a manner similar to Scanmail. In addition, a user can search a web page, including both text and speech attachments, via speech. Georgia Tech has some unique resources to do these tasks. (Thad)

store and forward google

In some rural areas, Internet service is store and forward. For example, in parts of rural India, the bus to the city has a wireless hub and computer on it. In the city it picks up the most recent email. As it drives through villages, it downloads the e-mail to village kiosks and uploads any e-mail waiting to go back to the city. (e-mail and many services were designed to do this from the beginning - see uucp, for example). Google searches can also be done this way. The user types in a search topic, the system downloads the top 20 hits, caches the related web pages, and searches on those searches, and caches those pages, etc. The data is delivered to the user's system and, for a while, it seems to the user that the search is happening live (until the cache is exhausted). There is a current sourceforge project on this concept. Determine a better way of caching the search (using other people's search history) for use with mobile devices.

Data driven sculpture

Using face recognition techniques, create a method for designing 3d sculpture that constructs a person's face using parts of flowers (or dog faces or cat faces or other people). Make a system that takes in a face an outputs a file for the 3d printer. (Thad Starner)

Electronic Wallpaper

Create wallpaper that changes colors, lights up, or senses the user's hands using conductive ink, thermochromic ink, electronic textiles, or however you can figure out to make it feasbile. Note there already are some commercial products in this space.



Augmenting toys to teach sign language. See the page for more info.


Daniel Ashbrook

This project involves making smarter hybrid cars. Hybrids currently have to make decisions about how much battery power and how much engine power to use while driving based on very little information. By taking GPS data to determine, for example, that the car is approaching a steep decline, we could tell the hybrid to go ahead and use up its battery power, as it will be able to charge it soon. Thus, we're more efficient.

DPS Watch

Daniel Ashbrook

Create a wristwatch interface that listens to the user's conversations and provides useful information. For example, “Let me see if I am available Monday” causes the wristwatch to display the user's schedule for Monday.

== Circular touchscreen wristwatch interfaces= Daniel Ashbrook

Investigate how to make interfaces for round touchscreen watches.

mini-QWERTY Tactile Study

James Clawson Keywords: mobile software development, user study, publication

For this project you will do some minor software development/integration and run a user study. We're investigating the impact of tactile feedback on participants' ability to input text on mini-qwerty keyboard equipped mobile phones. In essence we are going to figure out how quickly and accurately people can type on an iPhone and compare those rates and accuracies to typing on a Blackberry.

Driving + typing

James Clawson Keywords: mobile software development, auto software development, user study, publication

Integrate OBD2 data with typing input to quantify how much worse we drive when we type, and how much worse we type when we drive.

Driving Telemetry

James Clawson Keywords: sensors, auto software development?, user study, publication

Augment a car with sensors to capture in-vehicle use of mobile devices.


James Clawson Keywords: cognitive/user modeling, publication

Investigate people's decision-making process to determine how people choose to use or not use a partictlular user interface/mobile device.

Screen Focus Study

James Clawson Keywords: software design/dev, study design, user study, publication

Differences in notes taken when presentation is given on a projection screen versus a blackberry screen

Biking Buffercam

James Clawson Keywords: hardware/software hacking, building new hardware, power management

Bike-mounted video loop for post-wreck finger-pointing.

Gloved Interface Use

James Clawson Keywords: software dev, study design, user study, publication

Apply UCD principles to design, prototype, and evaluate mobile UIs for gloved use.

A wearable two-way tactile display with vibro- and electro-touch


  • Purpose: Designing and implementing a two-way tactile display with vibro-touch and electro-touch on a wearable interface. Each mode of touch (vibro- and electro-) should be overlapped in one surface with same resolution.
  • Motivation: One of the most critical concerns on multitasking with mobile devices is derived from serious visual distraction which is caused by the high demand on visual attention. Tactile displays provide alternative mode of output (alert) so that the user can manage limited attention with efficient way.
  • Previous works: EMMA (Electro-touch for mobile multitasking alert), Piano touch See more detail here
Haptic Gesture: Mobile gesture interfaces with tactile feedback


  • Purpose: Augmenting the input gesture of the mobile gesture interface (e.g., gesture watch, gesture pendant, sign glove) with tactile feedback to provide vision-free confirmation
  • Motivation: One of the advantages of the mobile gesture interface is its touch-free nature which enables interaction 1) with sensors even smaller than the users’ fingertip 2) when the users’ hands are contaminated. However, the remaining challenge is the lack of feedback when the input gesture is applied to the sensor. In this project, you will design and implement a wearable interface that synchronizes gesture input with tactile output which will eventually enable vision-free gesture interface.
Augment my Ride

Nirmal Patel

This project is an attempt to augment my car's standard gear shift with one that detects gestures made around it. You would build a new gear shift with embedded proximity sensors. These sensors would then be connected by bluetooth to an embedded computer that is hooked into my car's audio system for output. The combination of gesture (for input) and audio (for output) would be used to control many aspects of the entertainment system. The idea is to provide an interface for the driver that does not require taking eyes off the road. Contact for details.

Useful Skills: We can do this project in pretty much any language that has a nice Bluetooth stack. Python is great for this sort of thing. Some hardware hacking will be involved. It would also be nice to embed the sensors in a nice plastic case that is made using the 3D printer. We would like for it to look like a regular gear shift.


Scott Gilliland

This project is a continuation of a current one. The goal is to create a ball, roughly the size of an orange, that is able to sense pressure on it's surface, as well as detect its orientation. The core of the hardware has already been constructed, but needs more work, as well as software to do recognition.


Nirmal Patel / James Clawson

This project will involve building a highly synchronous location based message system. This project follows along an arc of CCG research that is focused on the collocated mobile experience. Unlike traditional messaging where a message is created entirely and then sent, this system will send every character entered to all of the people around the user. This will mean that the user will also see multiple continual streams of information and the primary purpose will involve exploring interfaces that easily support the understanding of the streams. We hope to build this as an iPhone application and then release it into the wild (AppStore) for data collection. Contact for details.

Useful Skills: HTML/JavaScript/CSS or Obj-C/CocoaTouch

Remind Me

Harley Hamilton

This will be a location-based and/or call based system that will send the user a message based on the location of other cell phones nearby or based on the number dialed by the user. Health messages may be a good focus. For example, if talking to a certain person raises the user’s blood pressure the message may tell the user to calm down before talking to that person. High “doing good” value for the project.


Harley Hamilton

A cell phone translator. What caller says is translated to language of listener. When listener responds it is translated to language of caller. Only needs to reside on one persons phone.


Harley Hamilton

A cell phone based lie detector. Work has been done on an automatic speech based lie detector to give you a start


Harley Hamilton

To curb impulsive responses and behavior a program would lock a user’s ability to respond to a text message for one to five minutes so the user can take time to consider what the appropriate response to the text message should be. Especially useful for teenagers to regulate impulsive behavior. A variation could monitor when a person is going to purchase music, ringtones, a car… over the internet. When the person startds to enter a credit card number the phone would time out for 5 minutes and ask “Is this purchase necessary?”

Dual Purpose Gesture (DPG) with Sign Gloves


  • Purpose: Designing and implementing a Sign Gloves for DPG. The glove facilitates real-time translation of the human readable gesture to the machine readable information.
  • Motivation: Gesture is widely used for face-to-face communication when the spoken modality is not available (e.g., ASL for Deaf people, Army’s hand & arm gesture). See here for quick examples. DPG enables two channels (co-present and remote) of communication. While you sign to the person of line of sight (human readable gesture), your hand shape and movement will be capture by the accelerometers and capacitive sensors on the gloves (machine readable gesture) and transmitted to the other party.
  • Previous work: initial Sign Glove, Dual Purpose Speech (DPS) See more detail here
Signal Glove

(basically, the same as DPG above) Using electronic textiles, make a combat glove for a game like America's Army that can be used to communicate with your remote soldiers using standard Army hand signals. Use tactile feedback to help the player know that he is performing the gesture correctly. (Thad Starner)

Sign Markers


TODO: add a description

Collaborative Mobile DDR or Rockband

Jiasheng/Thad/Nirmal Patel

Nirmal created a version of Dance Dance Revolution for a mobile phone which used our Bluetooth accelerometers for sensing the user's kicks. Extend the idea to having social gesture games on mobile phones. Make a team dancing game (e.g. country line dancing or music video) or a version of RockBand using sensors in today's mobile phones and our accelerometers, proximity sensors, and other devices.

Proximity Sensing Duplos

Scott Gilliland

We hope to create children's toys that can tell how they are being interacted with. One problem is telling when a toy is being handled by a kid - a perfect place to use capactive sensing. The challenge is to construct a toy that can sense when it it being held by a person, not just being bumped accidentally.

Awarehome Info

Brian Jones

The Aware Home provides access to an authentic home environment ripe for new application. Current efforts include aging-in-place and retirement communities, sustainability and energy conservation, home automation and indoor location sensing, and inter-household family communication. You can find out about these and other projects here:


Brian Jones

2link is a concept using 2 Ultramobile PCs for a communication device to encourage socialization in a retirement community environment. Developed in an Industrial Design class in Fall 07, we would like to produce a prototype of this concept for evaluation later this semester in different retirement communities and for eventual inclusion in the Aware Home demos.

Dementiacare Photoframe

Brian Jones

Similar to the concept of targeted advertising in retail stores, can a digital photo frame and backend be developed that could display photos targeted at the individual passing by or standing in front of it. From observation of people living in a dementia care facility where they share a common area, this seemed like one possible way to make the space feel like it is customized for them and possibly trigger some happy feelings.

Contacts Reminder

Brian Jones

This is probably quite simple, but something that I (and I am sure others) would greatly appreciate. Contacts that are extended wear sometimes rip or become irritable before their expected life, resulting in changing out one, but not the other. The date that you last changed contacts, so the date that you change them out gets out of sync. Designing a wrapper for a case that can be set to a specific change period (2weeks, 1 month, 3 months, etc) and where it can be reset when new contact use is started would be very useful.

Home VidChat

Jonathan Sheppard

This is a system which would center on an interactive home messageboard and whose main feature would be video chat. Although video chat is hardly something new, this system would employ portable vidchat rather like a cordless phone, designed for use around the house. Through the use of a transparent OLED display with a camera embedded behind it, eye contact could be maintained through two-way videochat. (An active image filter would subtract the display's interference from the image captured by the camera). In current mass-market systems, a participant can look at the display or the camera, not both - resulting in a lack of eye contact and the impression that the other person is not paying attention.

Possible opportunities here include building a working in-home portable vidchat unit, reevaluating the system from a more technical standpoint and experiment with working models and interfaces, and expanding on the system with mobile devices or a device targeted at the elderly.

Super Remote: TV Remote using a wiimote

Shantanu Talapatra

This project will use a wiimote as a sensor and an IR wand with 2 buttons as the TV remote. An open source media pc software package will be used as the software to control the TV, tentatively we will be using MythTV. The idea is to be able to drag through the tv guide like a google map, flick through it quickly, and pick exact shows. Also, while watching TV to be able to flick through channels and implement other cool ways of controlling the TV's ui. Basically we want to make this like a touch screen from a distance.

Leg gestures

Daniel Ashbrook

DEKA (of Segway fame) has a prosthetic robotic arm. They need to control it in some method, and are interested in leg gestures. This project would involve creating a control system with one or several leg or angle mounted accelerometers.

Smart Safety Kit - for workers working in hazardous places

Kedar Toraskar, Sudnya Padalikar, Matt Luongo {ktoraskar, sudnya.padalikar, mluongo} We are looking for one more member.


Workers working on Electrical Lines, Sewers and other such places work in hazardous environments. They need to work with different types of tools in environments such as closed spaces and suspended on poles. They also need to carry along and use bulky communication equipment while working. This can pose a big safety hazard. Through our Smart Safety Kit, we intend to address this issue. The project is to create a smart kit which involves effective visual and verbal communication and provides ease of information access and sharing. This kit promotes safety in usage of communication equipment by means of simple commands via gestures.

Description: The system will address the following issues: 1. Visual and verbal Communication 2. Availability of Training Manuals 3. Easy Gesture based control of the entire system Main Hardware: Glass Mounted Display, Camera, Bluetooth Earpiece, Gesture Recognition Gloves, Simple computing controls embedded in the jacket - like scrolling, selection etc., a bar code reader to note the parts replaced/modified while on the job, a belt mounted computer (this computer can even be kept in the service van) * The worker will wear the Glass Mounted Display and the camera will be mounted on the helmet. This arrangement can facilitate 2 - way visual communication between the worker and the control station. This can help in real time problem analysis. * The worker can view training manuals on the screen mounted on the helmet. * The worker will wear a bluetooth earpiece which will be connected to the communication device (radio, cell phone) * The worker will wear the Gesture Glove - which will be used to activate the devices by using simple signs * The worker will wear the safety jacket which will have within it embedded controls for scrolling through the menus and selection on the screen * The safety jacket will also have within it an embedded bar code reader for keeping track of the equipment that is begin removed/added * The safety jacket will also have a compartment with a battery to power all the devices. It will also have a slot to store a small computer (else it can also be kept in the service van and there can be wireless communication (wifi, bluetooth), between the computer and the devices.

Advantages: The worker is in a high risk position and his hands and feet are occupied in using the tools and equipments and balancing etc. By having this smart jacket the worker can keep his hands free from bulky radio sets and other equipment like manuals etc.

Since the scope of this project is quite big, we plan to divide it into 2 sub projects for the class.

classes/muc/projectideas.txt · Last modified: 2010/09/07 12:56 by abowd = chi`s home Creative Commons License Valid CSS Driven by DokuWiki do yourself a favour and use a real browser - get firefox!! Recent changes RSS feed Valid XHTML 1.0