Team Summaries - 2016


Intel-Cornell Cup Finalists

ASU IOT Education, Arizona State University

Visual programming languages have been
fundamental computer science concepts and paradigms without bogging them down
with syntax. The exemplar visual programming language for this purpose is
undoubtedly Microsoft VPL [1], which is a robust, feature-rich visual
programming language which supports a large number of peripheral hardware and
technologies. By using Microsoft VPL to interface with and control robots,
students in introductory engineering courses could, with little to no
programming experience, begin programming robots. Recently, however, Microsoft
dropped support for Microsoft VPL [4], leaving a void in the visual programming
language market. ASU-VPL is a proposed successor to Microsoft VPL that seeks to
act as a drop-in replacement for Microsoft VPL in introductory engineering
courses. The combination of ASU-VPL with custom built Intel Edison-driven
robots looks to replace the well-established duo of Microsoft VPL and Lego
NXT/EV3 robots. Challenges involved with extending ASU-VPL with additional
features, developing on-Edison software capable of interfacing with ASU-VPL,
and assembling an Edison-controlled robot were all encountered and addressed
during the course of this project. New features such as custom event callbacks
and web service support extend an already robust visual programming language in
ASU-VPL and allow instructors to intuitively explain these complex paradigms to
first-year students that have very little experience. Bash configuration
scripts for the Edison's Yocto Linux operating system make deployability easy,
and ensure that the ASU-VPL system can be scaled to a large number of students.
Support for ultrasonic and touch sensors enables users to invoke simple
wall-following algorithms to move a robot through a maze, a staple of the ASU
FSE 100 engineering course. 3D printed mounts and chassis keep hardware cost
down and put an emphasis on scalability.

 

Baymax, University of Massachusetts, Lowell

The need for medical assistive technology has
become increasingly prevalent in the last decade, fueled by the high cost and
difficulty finding trusted in-home nursing services. More recently, growth in
robotic technology has been expedited by intelligent computer vision, latest
battery technology, and energy efficient parts. However, little has been done
to bridge the field of robotics with medical assistance in a consumer-grade and
economical manner. In this paper, we propose a small scale and affordable
Unmanned Aerial Vehicle (referred to as a Baymax) equipped with an Intel
RealSense depth sensing camera to predict, detect, and even prevent impending
risks against a non-critically injured stroke rehabilitating patient. Baymax
will incorporate the aforementioned optimizations in the robotics field to
provide the level of care comparable to an in-home registered nurse within the
affordability of common household medical devices. In particular, this paper
will discuss the process and materials required to build a quad-rotor drone for
under $1000 that is capable of predicting, detecting, and even preventing an
imminent stroke, and conclude by briefly discussing the potential use of this
drone for other medical services offered by a professional nurse.


Biogaming Shoe, Johns Hopkins University

Johns Hopkins



Individuals with upper limb deficiencies such
as amputations often have difficulties handling computer keyboards and mouses.
While there are alternative methods to use these devices, they usually extend
only to slow typing and basic use. We propose a system that allows for people
with upper limb disabilities to retain additional control over the computer,
allowing for participation in recreational activities, such as video games.

Current solutions and advocacy for the control of videogames by individuals with
different types of disabilities is mainly provided by The AbleGamer Foundation.
To quote the words of founder Mark Bartlet, “I believe that there is nothing
more powerful for people with disabilities than the freedom that only
videogames can provide. It is an art form that allows us to run, jump, and be
whatever we want to be.”

We propose the design of a simple, but very targeted device for helping the user move
forward, backward, left and right along with jumping and crouching with the use
of their feet. The design will consist of force sensors and a Intel
microprocessor that utilizes an algorithm to determine these various commands.
This will communicate the desired inputs to the computer using a wireless
interface. It will feature a user interface for customization and also a sleek
simple mechanical design for maximum user comfort and ease of use.



BreakerBot, Boston University

Industrial 13.8kV, 800 pound circuit breakers at
Consolidated Edison electrical substations are frequently removed from their
housings for maintenance, replacement, or testing purposes. The current process
requires human operators to physically insert and extract the breakers from the
cabinets. Arc flashes (high temperature and explosive reactions) can occur if
faults exist in the breaker; these flashes can manifest themselves when the
breakers are removed from the cabinet, putting the human operators within the
arc zone at danger of burns, electrocution and concussive injury without the
proper safety equipment. Further maneuvering of the breakers is also difficult
and labor-intensive due to their immense weight and design. The goal of
BreakerBot is simple: remove the human worker from the equation. With
BreakerBot, the worker is replaced by a semi-autonomous, remote-controlled
robot. This robot is able to intelligently align itself with the breaker
cabinet and extract or insert circuit breakers as needed, all while keeping
people out of the arc zone and away from danger. Furthermore, the robot can
carry circuit breakers and provide a platform for the breakers to assist in
maneuvering around the substation floor, further reducing the workload and strain
on operators.



DAPS, Worcester Polytechnic Institute

DAPS

The goal of our project is to create an above-knee
lower-limb prosthetic socket that is able to adjust the interior socket volume
depending on the current activity of the user to optimize comfort, residual
limb health, and lower limb control. Smart variable geometry sockets have shown
promise in addressing volume fluctuations by lining the socket with
fluid-filled bladders to increase or decrease socket volume dependent on limb
size. Compression/Release Stabilized (CRS) sockets have shown that smaller
areas of higher pressure produce a better fit and increased user safety. We
will utilize key aspects of both prosthetic technologies to create a socket
that is more supportive, responsive, and better fitting. The proposed device
will actively monitor user activity with EMG sensors and limb volume with
pressure sensors. Sensor information will feed into a microcontroller on a
custom PCB. This microcontroller will control linear drivers attached to syringes.
These syringes will be directly attached to the bladders. Bladder adjustment is
fully automated except for some safety features. There are few commercial
sockets with bladders due difficulty of retaining control in an emergency.
During any emergencies, our manual fluid release valve will retain user control
until a safe position is reached.


Scalable Robotic Exoskeleton and Tablet Gaming Suite for Hand Function Rehabilitation, Georgia Institute of Technology

 ... 2

 There exist many debilitating neurological
disorders that reduce motor function; a few common examples of these disorders
are cerebral palsy and stroke. However, patients who participate in physical
therapy sessions, a common treatment for these conditions, may be able to
regain or retain some motor function. Unfortunately, many physical therapy
patients become bored with therapeutic exercises and, as a consequence, reduce
their participation, which lengthens recovery time. This project will extend
previous efforts of robotic assisted wrist physical therapy by increasing
engagement and the population who could benefit from this technology. Our team
created a scalable, light-weight robotic exoskeleton that functions as a video
game controller along with a tablet gaming suite that facilitates therapeutic exercises.
Our system has been validated with healthy subjects and has shown that
engagement increased by a factor of four as compared with traditional
exercises. Preliminary results also showed that exercise sessions are more
productive when using a fitted, lighter weight exoskeleton.

 

E-FIRE, Boston University

efire

In past efforts to measure electric fields in
near Earth space, large, heavy, and energy-intensive tools have been deployed
on satellites. This leads to exorbitant costs in both materials and fuel. Our
project, Energetic Field Instrument using Radiated Electrons, or E-FIRE,
proposes a new method of taking electric field measurements in a manner
suitable for deployment on small satellites using electron drift. We will
design a scaled-down, proof-of-concept model that will serve as the building
block for a larger, more complete design. Using a radioactive Bismuth-207
source to emit electrons, our instrument measures the energy of radiated electrons
using a diode array and a mixed-signal custom IC known as MIROC2. A user
interface will be used to display information such as diode hit detection and
electron energy. By using a passive source and small footprint, our system will
be substantially easier and cheaper to deploy for near Earth missions.
Exhibiting this functionality is crucial to motivate the creation of a
full-scale system, which would involve eight MIROC2 chips, 128 diode
detectors, and be capable of accurate electric field measurement in space.



Eggplant, Rensselaer Polytechnic Institute

eggplant

Wireless IoT Laser Projector: Laser-based
projector that can be remotely controlled and displayed from from any
computer/internet connected device. (Notable features include wireless display
and remote computer control.)

 

F1/10 Autonomous Racing, University of Pennsylvania

Driverless Cars have been around for quite some time,
however they tend to operate at slow speeds when in complex environments,
limiting their utility in high-tempo applications. We are developing a platform
that will serve as a base for competitions on autonomous race cars. We are working
on 1/10-scale model F-1 race car (with 10 times the fun!) capable of travelling
at max speed of 40 miles/hour. The algorithms developed will surround real time
perception, control,planning and response to fast changing environments. The
performance will be evaluated on the basis of speed, agility and power
efficiency - the three core challenges that even the full size driverless cars
are constrained by. The car will be equipped with several sensors such as
LiDAR, IMU, and Camera.The platform and the algorithms developed as a part of
this project will be made accessible to anyone willing to contribute to the
work with an aim to make highspeed driverless cars a reality.



Girodicer (Ice Dam Drone), Boston University

The goal of this project is to create a drone capable of
autonomously scanning a slanted residential roof to detect and alleviate ice
dams. The drone will utilize thermal image processing to detect “warm spots”
where ice dams are likely to form downwards from that point. In addition, the
drone will take pictures of the perimeter of a roof, process the images to look
for icicles (a sign of a forming ice dam), and make note of the problem area.
The drone will be capable of distributing calcium chloride (a non-corrosive
salt) within a porous fabric sleeve onto an ice dam to prevent further damage
to the roof. A smartphone app will serve as the interface for the user,
displaying information such as rooftop images and battery life. The drone
provides a safer and cheaper alternative to removing ice dams compared to the
dangers of manual removal or expensive roof heating solutions.


HabTech, University of Pennsylvania

Strokes are the leading cause of disability in the United

States. Each year approximately 800,000 people will suffer from a stroke; 80%
of whom do not regain full motor control and 35% will ail from the effects of
muscle spasticity.To remedy this, the HabTech Upper Body & Arm
motor-driven system that will allow patients with limited range of motion, due
to nerve cell damage, to regain motor control. HUBART will be focused around
the elbow joint and provide an isokinetic resistance to the patient until the
patient can no longer complete the range of motion. This means that the HUBART
system will automatically adjust the resistance, allowing the patient to
perform the range of motion exercise at a constant speed. Once a range of
motion limit has been reached, HUBART will begin to provide assistive torque
allowing the patient to achieve a more complete range of motion. This has been
proven to be a successful treatment to help reconnect neural pathways among
stroke patients. HUBART will be an at-home, space-efficient device meant to
promote continued physical therapy after patients leave the hospital. 

 


HAMR, University of Pennsylvania

HAMR is a low-cost holonomic robot platform based on a
unique drive mechanism that is simpler and more robust on uneven terrain than
conventional omnidirectional systems, such as powered casters and omniwheels.
By providing maneuverability for other robotic systems like manipulators or
robot-human interfaces, HAMR can enable research and lower the cost barrier for
consumer robotics.


 

HERO Glove, Worcester Polytechnic Institute

6

The Haptic Exoskeletal Robot Operator (HERO)

Glove is a system that is used to teleoperate robot manipulators whilst
providing realistic, accurate haptic feedback. Our system consists of two
parts: the master glove and the slave module. The master glove achieves
intuitive tactile and force feedback through use of soft pneumatic actuators at
the finger tips and joints, that stiffen correspondingly to the interfaced
manipulator's interactions with its environment. Accurate finger joint angle
measurements are obtained through curvature sensors that were developed by the
WPI Soft Robotics Lab. Positional control is obtained through a 9-DOF IMU. An
electronic control module allows for an easy add-on via USB to existing
manipulators. The module receives the glove's 3D position, and passes it to the
robot software for position control. Tactile and joint sensors will be stuck on
the manipulator connected to the electronic control module, which will then
transmit this data back to glove. To achieve predictable bilateral
communication performance, the wireless system is optimized for high speed and
low latency with error handling protocols. A concept of operation is having a
user control a manipulator to open a door remotely.


Mooove, Boston University

00

Mooove is a suite of sensing devices that work together for
the ultimate goal of making lives better for cows and their owners. Anklets
worn by the cows exchange information with one another, creating an efficient
mesh network. Information such as temperature, sound, and location are tracked
and sent to the cloud. Through the use of an interactive web UI, users can
easily access the processed data of each of their cows. What’s unique about
Mooove compared to existing products is that not only does this keep the
location and analyze the health of the animals, it also utilizes a complex
algorithm to figure out what the animals’ social lives are like. Studies have
shown that cows have best friends, and will get stressed if they are separated.
Stress in turn may lead to lower milk production or health issues. By using
Mooove, cow owners can take tracking to the next level, allowing them to
increase increase production and the quality of life for their cows.

SAR Lynx (HazardnAssessment Fleet), University of Colorado Denver

Manmade and natural disasters, civil unrest and violence,

call into light the increasing need for technological innovation in Urban
Search and Rescue (USAR). Before USAR personnel can enter a site to rescue
victims, they must have critical information regarding (a) number, location,
and condition of survivors, (b) maps of the scene, and (c) environmental
conditions. This information is assembled by USAR operators into a “hazard
assessment” which will guide USAR operations on that particular scene. The
creation of this hazard assessment must be done, and cannot be neglected or
shortened even if it is suspected that prompt rescue could save more lives. The
Hazard Assessment Fleet (HAF) is a cooperative team of robots, deployable by
first responders, that will autonomously provide a real-time 3D environmental
map, live video, two-way communication, and environmental data. In this manner,
the hazard assessment will be complete upon the USAR team’s arrival,
significantly shortening the time from arrival on scene to active search and
rescue. The robots will then also team with USAR personnel to monitor the
scene, provide real-time updates, and assist in search operations, as directed
by USAR personnel.


SCUBAssist, University of Pennsylvania

link

SCUBAssist is a semi-autonomous underwater
vehicle that aims to be the ultimate SCUBA diving assistant by performing three
main functions: scouting dive sites; autonomously following and recording video
of a SCUBA Diver; and serving as an underwater navigation aid. Currently there
is no single or economical solution that that assists divers with these tasks.
To scout dive sites, the vehicle will be tethered to a surface controller that
allows divers to view live underwater video. Divers can then determine dive
conditions and whether the site is worth exploring. Once a dive site is
selected, SCUBAssist can be un-tethered and used as an semi-autonomous
videographer. The vehicle will triangulate the diver’s location and follow the
diver via an acoustic beacon attached to the diver's air tank. SCUBAssist will
be compatible with a GoPro camera and will use a gimbal to stabilize the
recorded video. A wrist controller will allow the diver to both control the
vehicle’s relative position and camera. SCUBAssist will also be a navigational
aid by referencing a pinger on the dive boat or start location to lead the
diver along a route or back to the start location.


SHAF team, Portland State University

Having a device that can help individual
maximizing the benefit gain from doing exercise and minimizing the probability
of doing harmful exercise is very helpful and important in our modern life. Our
team plans on creating a device that can be attached to parts of a user’s body
to test the muscles' activity level. Since doing a specific exercise can cause
specific muscles on a human body act intensively, by looking at the
measurement, users can know if they are performing the exercise correctly or
not and if that exercise activates the correct muscle that they want to
activate. In addition, if users are in a bad physical health condition and
there are some muscles they are not allowed to be activate, the device can
alert the users so that they can avoid doing harmful exercise. In order to
alert users, LCD display and audio feedback would help the users know when
they’ve done exercise correctly and if the exercise is right for them.


Suit Up!, Worcester Polytechnic Institute

Suit Up!

Mastering a new physical skill currently involves visual or

verbal instruction and hours of practice; however, physical feedback could
expedite this process. Our goal is to create a system that incorporates motion
capture and real-time haptic feedback to streamline the learning process and
give individuals more autonomy in physical training. To achieve this, we will
develop Wearable Action Guidance (WAG) bands, which will be worn at strategic
points along the hands, arms, and torso. Each of these bands will communicate
its position and orientation to a central computer for processing and recording
to a motion capture file. This file can be edited and then uploaded to a
centralized library for further sharing. Later, another individual can choose
an existing motion capture file and use the WAG bands to receive haptic
feedback indicating deviations between his or her motion and the selected
motion file. The software accompanying this device will include playback
controls and basic editing capabilities, as well as a visual display to
simulate recorded motions. We see our system being used as a teaching tool for
those wanting to learn new physical skills.


TED – Translating Educational Device, Boston University

Modern communications have made the world a smaller place.
Geography is not a limit to contacting people living in even the most remote
places. One barrier that still exists, however, is the language barrier. T.E.D.
(Translating Educational Device) seeks to address this communication barrier by
promoting and fostering a generation of multilingual children. TED is a teddy
bear toy with the primary function of teaching another language to children
ages 3-5, the age at which children are most adept to becoming multilingual.
TED is composed of custom hardware for sensing and imaging, natural language
processing software for translating, and complete integration with the cloud.
Additionally, a smartphone application allows the parent to change settings and
view metrics tracking the child’s progress. Ultimately, TED is intended to
teach children ages 3-5 another language by mirroring multilingual
communication usually only provided in families with bilingual parents or
guardians.


Terminator Arm, Howard University

TA ...

This project seeks to create a relatively
inexpensive 3D printed prosthetic arm that is controlled by electric action
signals from the brain. The process of capturing these signals and translating
them to hand motion is to be carried out by a Myo armband (using
electromyography) and an Intel microcontroller board. We seek to create a
prosthetic arm that can mimic basic hand gestures including finger extension
and contraction, fist clench and palm spread. We also wish to design a
prosthetic that is sensitive to unfavorable heat conditions and alerts the user
of conditions that could potentially damage the prosthetic arm. Furthermore, we
wish to build a prosthetic arm that gives a heads-up to its user about a low
battery and is also easily removable for maintenance/comfort when sleeping etc.
Finally, we wish to build a prosthetic arm with a higher degree of
functionality than regular prosthetic arms in the market.


The Pyromaniacs, Purdue University

pyro

To commemorate the historic bicentennial

anniversary of the statehood of Indiana, Purdue University has been contracted
to develop a Smart Torch for the event. Our team is responsible for the
development of a robust electronics system within the torch. The state-defined
goals for the electronic functionality of the torch require inclusion of a
flame control and safety mechanism and an ability to take pictures and videos,
using a simple and elegant user-interface. The torch should also be able to
communicate real time data about traversed location and send high definition
picture/video across the World Wide Web to a server dedicated to the project,
which will collect, process, and store the data in a central location. Our
design implements the idea of the Internet of Things (IoT) by utilizing the
Intel Edison computing module as the core controller of the torch. This design
is unique, as never before has a torch been developed with “smart” features.
Additionally, due to the statewide exposure of the project, this project will have
the opportunity to excite and inspire future generations of engineers and
scientists throughout the state of Indiana.


Tufts Robotics, Tufts University

...

In the wake of an earthquake, the goal of first responders
is to find and secure as many people as possible, as quickly as possible. Our
goal is to improve the effectiveness of rescue teams at rescuing earthquake
survivors. We will accomplish this by designing and building an autonomous
robot that will aid in the search process by finding and marking the locations
of survivors in the area. Then rescue teams will be alerted of victims’
locations so that they can render help. In this way, the robot expands the possible
search radius without any additional effort from the human response team.
Robots that aid in search and rescue have been developed before, but none to
date have been autonomous. Preexisting solutions require the attention of a
human driver who might otherwise be aiding in the rescue effort. The main
benefit of these kinds of robots is that they allow the exploration of hard to
reach areas, whereas our robot is primarily concerned with increasing the
efficiency of the search for survivors.


UCD UAV Team, University of California, Davis

Wireless sensors and sensor networks show a
great potential in agricultural and environmental applications where persistent
and pervasive sensing and monitoring are needed. Although technology
development in this area has resulted in ever smaller and more capable sensors
and processors, the limited battery life of unattended sensors remain an open
challenge to practical implementation of wireless sensor networks over large
areas. Adding to the problem is the requirement to transmit and received the
sensed data over long distance. Autonomous and unmanned aerial vehicles (UAVs)
provide a promising solution to the above challenges. Low flying UAVs can be
used to read data from a remote sensor within short distance therefore saving
precious battery energy on the sensor. Our team intends to integrate a
quadcopter with a data receiving system. The quadcopter will also have improved
situational awareness and collision avoidance through on-board distance
sensors.


Team Vispi, University of Pittsburgh

We offer a completely autonomous solution to the
logistic challenges of parcel delivery in a residential setting. Our project
seeks to minimize delivery time as well as eliminate human operator
involvement. We seek to accomplish this through a fleet of drones that are
deployed from a self-driving vehicle. The drones return to, communicate with,
and get resupplied by the central vehicle, analogous to planes on an aircraft
carrier. Similar solutions under development use only one drone, rely on a
human pilot to control the drone either from a distant communication station or
the central vehicle. By enabling the drone to autonomously determine its
relative location and subsequent flight paths, we eliminate the need for a
pilot. Combined with an autonomous vehicle capable of navigating through city
streets and finding optimized delivery sequences, we arrive at an efficient
delivery system requiring no manual human operation. Such an efficient system
would increase the number of deliveries per hour by an order of magnitude,
leading to the possibility of multiple runs to a residence in the same day.
Once our proposed system becomes fully operational, the cost and time of
deliveries will also be reduced to a fraction of the current system.

End of 2016 Finalists


EnTech, University of Houston

EnTech

A self-sustained, water purification system that can be
cheaply made and effective in getting rid of bacteria and other unwanted
viruses.

 

Green C3, Pennsylvania State University

In an effort to create a greener environment,
our project addresses the problem of recycling garbage found in areas after
events such as music festivals, football games, and other popular assemblies.
The recycling rates need much improvement; many environmental organizations
estimate only around 27% of plastic bottles are recycled. The goal of our
project is to build a low cost prototype of an autonomous multi-system composed
of a drone, a ground station, and rovers that work together to clean a
designated area. Specifically, the drone will provide aerial images that will
be used by the ground station for intelligent coordination of the rovers. The
rovers will be equipped with the ability to identify and collect waste.
Furthermore, the ground station will provide image processing computation,
command control, and communication. Our unique approach, inspired by the
division of strategic and tactical systems often found in the military,
overcomes the limitations of each individual systems by exploiting its
strengths.

 

GUAVAS, University of Florida

Autonomous Autopilot System for a quad-copter
using Waypoint.

 

 Hermes, University of Pittsburgh

The purpose of this project is to create an autonomous
aerial drone capable of evaluating an environment to determine an optimal
location to deliver supplies. This device fulfills the needs of markets ranging
from commercial delivery to search and rescue (SAR) teams. Additional
functionality can also be added to specialize the product for certain markets.
One proposed piece of additional functionality is the ability to autonomously
detect human life and mark the location of the detection on a map for further
action, which would be especially valuable to SAR teams.

 


K. Dots, University of Pennsylvania

kdots

Braille is a tool for empowerment. Text-to-speech devices
are incredible aides to the visually impaired, but Braille literacy is crucial
for independence and success. Just as students with sight struggle to learn
solely through spoken word, visually impaired students need tactile resources
to progress academically. Currently, only 10% of visually impaired students are
being taught to read Braille, holding the visually impaired community back from
educational advancement. To encourage educators and young learners to tackle
these challenges early, our team proposes a low-cost, kid friendly e-reader.
Unlike cost-preventative solutions like piezoelectric crystals,
electro-mechanical actuation will allow for high-accuracy and durability at an
accessible price point. While refresh rates may not hold to the standards of an
advanced braille reader, young learners simply need a device that encourages
their education through displays of letters and vocabulary. By minimizing total
parts, the device can stay affordable for students of all financial means.
Paired with a qualified educator, a reliable low character display with a
moderate refresh rate is all a child needs to start his or her path to
educational independence. We plan to enable these students by giving them
access to their first tool towards a life of equality.
 

 

NoLeak, University of Illinois at Urbana-Champaign

Water leak detector utilizes cutting-edge audio signal
processing skills to listen to the water pipe in fire hydrant and predict the
existence/severity of the water leak.
 


OSU-ASCC, Oklahoma State University

The excessive use of authority by Police has created
nationwide controversies in recent years and has become a root cause for
violence and civil disorder in some states. To the rescue, body-worn cameras
are being integrated in the day to day operation of Police forces among other
measures. While this appears to be promising, there remains a number of
challenges to fully utilize the technology. These include, privacy of citizens
being recorded, easier querying of video segments and early warning of use of
force. This project is aimed at building a smart body-worn camera which could
solve the above problems and maximize the efficiency of the system. Our
body-worn camera can be operated using voice commands and has the features: to
automatically start recording videos upon detecting emergency situations, to
inform civilians that they are being recorded, to label activities from the
video segments and to localize and map an emergency scene and the officer. The
system will be built using intel Edison board, a stereo camera, microphone and
speaker, motion sensor and a GPS unit. The real time detection algorithms will
be implemented on the Edison and labeling of video segments and storage will be
implemented using cloud servers.

 

 

 R2DR, University of Massachusetts, Lowell

C3PO

Falls are one of the leading causes of injury
and death for Americans over the age of 65. According to the Centers for
Disease Control and Prevention, 2.5 million people over the age of 65 are
treated for fall related injuries each year. While there are many products for
quick emergency response available such as Life Alert, Rescue Alert, and Quick
Response Medical Alarm, these products do nothing to detect or predict a fall.
However, recent advances in the fields of robotics and machine vision allow the
detection of abnormal movements so that emergency services can be contacted
before a fall occurs. The team proposes to accomplish this by developing a next
generation, intelligent medical robotic system called R2DR - Robot to Detect
and React to gait degradation. The team plans to utilize the iRobot Create 2 in
conjunction with the Intel RealSense R200 3D camera to track and monitor the
patient in their home. Data processing of patient movement will be handled on
board the robot via the Intel NUC computer. By analyzing gait movement in real
time, the R2DR will be able to accurately predict an imminent fall, and
immediately contact medical or emergency services.


Roaming Spirit, Worcester Polytechnic Institute

With the rampant popularity of robots, more industries and individuals alike are entering the field of
robotics. However, with a lack of funds and high level engineering
knowledge/experience, many people will not be able to give life to robotic
creations. Since localization and navigation are key aspects of an indoor
mobile robot, we propose to create a sensor suite that abstracts the
hardware/software needed for indoor localization and navigation away from the
developer.The sensor suite must be able to take in external information from its 
sensors, kinematic equations describing the motion of the robot, and a mission 
for the robot. This mission maybe to explore aimlessly, or to reach a certain point 
on a map. From the information, the suite must then process sensor data in 
context to the mission,and produce a series of moment commands for the 
robot.  Unlike other, similar solutions, this sensor suite must be able to process
 all gathered information on board, rather than needing to stream information 
to an external computer. This project will improve the capabilities of indoor
 mobile robotics in a lightweight, cost effective manner such that any development
 team - corporate or maker - can innovate.
Robot Mobility for Children, Georgia Institute of Technology
Toddlers with severe motor disabilities who are
unable to develop the means to explore their surroundings may also be at risk
for other delays or even impairments as a result of missing early learning
experiences. Wheelchairs are not a viable solution for these young children as
they are extremely expensive and insurance companies make it difficult for
young children to be approved. However, research has shown that the navigation
of a baby's environment is crucial for brain development. Many times, children
with motor impairments develop other functions more slowly (talking, learning,
etc.) as a result of their lack of mobility. To fill this need, the team will
be modifying commercially available, electric, ride-in toy cars with new
control schemes and safety features that meet the special conditions of
children with motor function disabilities. These control schemes will permit
children with severe motor impairments to become mobile and explore their
surroundings.

 

Smoke Sentinel, University of Rochester

A drone that will autonomously monitor an area
for signs of a wildfire.

    

Thunder Dawgs, Southern Illinois University

In order to prevent potential power outages and to maintain
expensive equipment, electrical utility companies spend millions of dollars
every year on the dangerous and time-intensive labor needed to inspect high
voltage power lines; however, with the proliferation of low cost, high quality
drones and the miniaturization of both computer and sensor systems, there is
significant opportunity for a market of products that will automate this
process. The Thunder Dawgs seek to examine how modern automation technologies
can be applied to power line integrity checks in the process of maintaining
employee safety, minimizing cost, and maximizing efficiency. The proposed
solution seeks to provide a configurable platform that can be mounted to a
drone that includes a suite of visual and thermal cameras as well as a microprocessor-based
development board that can be used in tandem to investigate power lines during
a pre-programmed mission. In order to guarantee results by the conclusion of
the project period, though, the scope of this proposed solution has been
limited to the development of a set of algorithms capable of identifying
regions of power line degradation in pre-captured images and to the
implementation of a small-scale test bench for the controlled testing of these
algorithms.

   

Virtual Realty, Boston University

A revamped and reimagined version of the classic
board game, Monopoly, where a single user is embedded in a virtual reality
board game against computer players. The game includes not only the physical
board which is traversed autonomously by robotic pieces, but also the virtual
board, where the player can virtually explore the game board and surroundings.
Both the physical board and the virtual game work in tandem to preserve all the
rules and components of the original Monopoly board game, while also giving the
user the opportunity to explore the virtual world developed on the Oculus Rift
through multiple vantage points. The virtual game will include visualization
for the player’s properties, money, houses/hotels, and current board position. There
will also be functions requiring user interface for rolling dice, reading
Chance and Community Chest cards, buying properties, and exploring the game
board through both a “bird’s-eye” view and a “walk-through” view.