The Tech For Social Good “Tech Development Track” provides funding support of $500 – $5,000 for projects that promote social good by supporting healthy, sustainable, prosperous, and equitable livelihoods in the United States and abroad.
2019-2020 Funded Projects
|
A Physical Activity Coach in English and Spanish for Individuals with Limited Health Literacy We are developing a physical activity chatbot delivered via smartphones, tailored to individuals with low health literacy and available in English and Spanish. Insufficient physical activity is one of the leading risk factors of death worldwide, but encouraging individuals to increase their physical activity is a daunting task. Interventions via smartphones can help people become more active, but they are not typically available in languages other than English. Further, their level of health literacy (knowledge about maintaining good physical and mental health) is often too high for low income or low education individuals. We use Artificial Intelligence (AI) to provide personalized physical activity recommendations and support, thus functioning as a ‘physical activity coach’. Doing so, we aim to improve health outcomes and work towards increasing (digital) health equity. Team:
|
|
|
Aerostry Aerostry is a sensor platform and analytics engine for tracking and predicting the movement of pollutants. Our mesh network of inexpensive and durable sensor nodes collects data and analyzes it using novel convection-diffusion models. In urban environments, Aerosty will serve as an important tool for public health and public awareness of air pollution, while in nature, Aerostry can be strategically deployed to allow early detection of fires in high-risk locations, like in areas prone to downed power lines or lightning strikes. Team:
|
|
|
Smile with Joy Smile with Joy is working to create more sustainable toothpaste packaging. Team:
|
2018-2019 Funded Projects![]()
|
81 Cents 81cents empowers women to more confidently, effectively negotiate new offers, raises, and promotions through personalized, crowdsourced offer reviews and coaching. Team:
|
|
|
I Regret to Inform You That Your Private Information Has Been Compromised In an era in which data privacy, surveillance, and powerful machine learning tools increasingly constrain American lives, understanding how the benefits and threats of these technologies are experienced by diverse communities becomes ever more urgent. Members of the UC Berkeley Interdisciplinary Research Group on Privacy are conducting a multistage, mixed-method study to investigate how data privacy, surveillance, and machine learning technologies impact diverse communities. Team:
|
|
|
xinampa: Tracking and Predicting Environmental Justice Hotspots in Latin America through Big Data and Applied Machine Learning xinampa is developing an open source platform to support collaborative tracking of environmental justice (EJ) across Latin America, beginning in Mexico. They will use their CITRIS Tech for Social Good award to pilot a SMS/WhatsApp EJ BOT in Mayan Communities across Yucatan to record images, videos, and text documenting social-environmental damage. These data will be used in tandem with official data, news articles, and community-level data to build a predictive algorithm of environmental justice across the region. Team:
|
|
|
Re-centering the Body in Technological Utopias What does it mean to have a human body in the future of technologically-driven innovation? Cutting-edge technologies promise to help us transcend human limitations to greater physical and mental capabilities. This project uses speculative design research to recenter the body in visions of such utopias. Team:
|
|
2017-2018 Funded Projects
|
|
Archer Flagship Public data is the foundation upon which investigations into issues like corruption and human trafficking are conducted. Archer Flagship, project of Archer Impact, seeks to streamline the collection and integration of public data, empowering investigators to hold illicit actors accountable.
|
|
EMPOWER EMPOWER is a modular, affordable, residential energy management system that leverages the predictive power of modern machine learning to simultaneously optimize local energy generation, storage, and flexible loads within a house or microgrid. It integrates weather, PV, and user behavior forecasts to increase efficiency and is retrofittable to existing homes. EMPOWER will be pilot tested at the UC Berkeley Tiny House In My Backyard (THIMBY).
|
|
|
GAIT GAIT is a technology-based, task-oriented feedback device that combines physical activity with cognitive stimulation to address maintenance of gait performance in community-dwelling older adults with non-neurological gait impairment. Users of GAIT will be able to track their gait over time, letting them know of changes or whether intervention with a professional is necessary.
|
|
|
Laika Laika, a walking four-legged robot, enables the transportation of vital supplies during disaster relief. Laika has a flexible, actuated spine that helps the robot position its feet on uneven terrain, enhancing balance and stability. Through CITRIS Tech for Social Good funding, Laika’s team will prototype a higher-fidelity actuated spine to enable greater locomotion over rough terrain.
|
|
|
MarHub MarHub’s migration management platform changes the way refugees access information and services throughout their journey. The team is building a chatbot that provides refugees with tailored information and connections to NGOs through an automated intake process. Long-term, MarHub’s crowdsourced platform will help refugees and humanitarian actors work together to provide and evaluate refugee-centered information and services.
|
|
|
Seeing Like an Algorithm Seeing Like an Algorithm provides an interactive look into facial analysis technologies’ applications, biases, and impacts on marginalized communities. By comparing several commercial facial analysis tools, the project aims to uncover biases in facial analysis algorithms and recommend means to improve the algorithms’ fairness.
|
2017-2018 Fellows’ Projects
Developing a Computer System to Augment SMS Helpline Counselor Training
Orianna DeMasi, PhD Candidate, EECS & Amit Talreja, Undergrad, EECS
Crisis and suicide prevention helplines seek to aid individuals in distress and difficult situations. These helplines often operate on tight budgets and thus finding the time and resources to train volunteers in counseling techniques can be challenging. However, properly training counselors is critical and counselors cannot learn “on the job”, as that could put vulnerable individuals at risk. As a result, a low risk-environment is needed to train counselors and give them the opportunity to practice crisis intervention strategies.
We developed an automated tutoring dialog system to augment the training of SMS helpline counselors. We aim to develop a system that can teach novice counselors conversational techniques and provide a low-risk environment for them to practice these strategies in a chat setting. An important feature of this system is providing feedback to the novice counselor so they may improve their counseling approach, as they practice. Developing such a system will improve the efficacy and efficiency of training of novice crisis counselors and thus help more individuals in need.
Gridwatch
Noah Klugman, PhD Student, EECS & Ben Ben Zour, Undergraduate Student, EECS
Despite the fact that access to reliable electric power is critical to quality of life and economic stability, utilities worldwide often depend solely on customer-generated reports to learn the location, duration and size of service interruptions. GridWatch is developing two new low-cost methods for monitoring distribution-level power outages and restorations.
The first tool, GridWatch, utilizes the power of the crowd by running outage detection software on the smartphones of electricity customers. The key insight of GridWatch is that unmodified smartphones can detect power outages by monitoring changes to the phone’s charging state, using existing on-phone sensors to identify when a change in charge state likely represents a power outage or restoration, and corroborate the automatically generated reports with other phones through cloud-based services.
The second tool, PowerWatch, is a fixed point sensor. Often, just knowing if the power is on or off can provide large amounts of value; PowerWatch can achieve this with significantly less complexity and cost, compared to traditional smart meters. The data stream from GridWatch and PowerWatch, combined with advances in machine learning, distributed data-stream processing, and complex network analysis, help inform immediate and longitudinal determinations about the state of the power grid, providing a significant improvement in the data available to utility companies, regulators, researchers, and ratepayers.
In partnership with local decision makers, we are conducting a deployment of PowerWatch and GridWatch in Accra, Ghana. This deployment will include over 350 PowerWatch sensors placed in households and 2000 GridWatch app downloads. This deployment will help us test the accuracy of the sensors and will allow us to ensure that the data stream produced can be incorporated into existing energy utility systems.
More info at grid.watch
Path to Data Democracy: Application of Blockchain Technology to Address Clinical Research Challenges
Madelena Ng, PhD Student, EECS & Zoe Husted, Undergrad, EECS
In the Precision Medicine era, researchers are turning to digital research studies for rapid study implementation and data collection. Digital research studies are heralded as the novel approach—harnessing mobile health technologies and internet ubiquity—to overhaul conventional standards, from recruitment and enrollment to engagement and retention. Researchers are optimistic that digital research studies will help them reach and engage larger and more diverse populations, in a cost-effective and efficient manner. However, racial and gender representation and participation in clinical research remain subpar.
The purpose of this study is to understand (1) how women participate in and experience a digital research study and (2) how blockchain technology can be used to address persisting participation challenges in digital research. The Global Women’s Health Outcome Study (GWHOS) is a prototype digital study that focuses on women’s reproductive health. Participants complete GWHOS research activities and donate digital health data electronically (i.e., exclusively via their smartphones) through the Bitmark app. Participants donate data to further understanding about reproductive health patterns (e.g., irregular menstruation) and help explore the concept of data ownership in digital research.
Street Story: Using Self-Reported Data to Address Demographic Disparities in Transportation Health and Safety Data
Kate Beck, Master’s Candidate, Dept. of City & Regional Planning and the School of Public Health & Nina Djukic, Undergrad, Environmental Health & Communications
Low-income groups, people with disabilities, seniors and racial minorities are at higher risks of being injured while walking and biking, however information on the safety needs of these groups is limited. Transportation planners and engineers use past transportation collisions that have been reported by police to allocate safety improvements. Police-reported data are limited in a number of ways; pedestrian and bicycle collisions are often not reported to police, specific demographic groups are less likely to report issues to the police, and collision data exclude locations that are unsafe but where a collision has not yet occurred.
Self-reporting, which is when individuals report issues directly to city agencies, is beginning to be used as a way of collecting data on transportation-related health and safety issues that augments traditional reporting mechanisms. Self-reporting has the potential to collect information from users who are under-represented in police reports and change the way safety resources are allocated. However, this method could under-represent specific groups who face barriers to accessing or friction to using self-reporting platforms, perpetuating existing issues in transportation safety datasets.
This project will analyze the limitations of using police-reported and self-reported data to make transportation safety decisions, and 2) design Street Story, a platform that provides agencies and advocates with the data necessary to better allocate transportation safety resources. The Street Story platform is being co-designed with community members, agencies, and advocates who represent transportation network users and decision makers to ensure the platform meets the needs of these groups. The platform is being developed for the California Office of Transportation Safety (OTS).
This project is affiliated with UC Berkeley’s Safe Transportation Research and Education Center (SafeTREC) and has previously been funded through the Center for Technology Society and Policy Fellowship program.
Youth Empowerment Project (YEP)
Yohee Choi, Master’s of Engineering Candidate, IEOR
YEP aims to enable youth to gain sustainable leadership skills through exposure to and education about technology social issues. This will ultimately empower them to become agents of change in their own community.
2016-2017 Funded Projects
