Agenda Item 2a: Good Systems – A UT Grand Challenge — original pdf

Backup

Good Systems and City of Austin Collaboration Overview December 9, 2020 City of Austin presenters (in order of appearance) University of Texas Presenters (in order of appearance) CHARLES PURMA IT Manager, CTM SARA SMITH IT Business Analyst, Senior, City of Austin TED LEHR IT Data Architect, CTM JENNIFER LYON GARDNER Deputy Vice President for Research ANDREA CHRISTELLE Good Systems Network Relationship Manager JUNFENG JIAO Good Systems Executive Team Chair Opening Remarks Charles Purma, III IT Manager, CTM Jennifer Lyon Gardner Deputy Vice President for Research Designing AI technologies that benefit society is our Grand Challenge... Vision Generate AI ethics protocols that are widely accepted, cited, and used. Mission Create human machine partnerships that address the needs and values of society. Mechanisms  Research Projects  Good Systems Network  Research Focus Areas FY21 Research Focus Areas C R I T I C AL S U RV E I L L AN C E I N Q U I RY We work with scholars and organizations to curate conversations and exhibitions that help people understand the social and ethical implications of surveillance. P U B L I C I N T E R E S T T E C H N O L O G Y We build teams that gather public, open, and accessible data, to integrate research with policy, journalism, and local activism. D I S I N F O R M AT I O N We support an interdisciplinary faculty research community that makes monthly research presentations, and sponsor special programming to advance the understanding of dis- and misinformation. M AC H I N E L E AR N I N G AN D R O B O T I C S We focus on how fairness and other ethical considerations are applied in machine learning and robotics. FAI R AN D T R AN S PAR E N T AI We work to create fair and transparent AI technologies that people can easily use and rely on. F U T U R E O F W O R K The way people work is changing. Our research explores new ways for people and AI to work together. R AC I AL J U S T I C E We serve as a resource for education, community engagement, and research at the intersections between racial justice and technology/AI. S M AR T C I T I E S We develop transformative technologies to achieve resiliency and sustainable growth in urban communities. Collaboration Scope We are bridging barriers between fundamental knowledge and real-world problems by connecting disciplines, techniques, and ways of thinking. • 7 Collaborative projects • 12 City of Austin Departments • 13 UT Departments and Schools Good Systems City of Austin Collaborative Projects • Austin AI Housing Analysis • • Inclusive and Trustworthy AI Governance Design Inspection of City Infrastructure via Peripheral Perception • ML4GIS: Developing and Evaluating Computer Vision Methods to Enhance Access to Geospatial Data in Large Historical Map Collections • Optimize EMS Responses during Extreme Events • Cameras, AI and Public Values in Smart Cities • Smart Cities Should Be Good Cities: AI, Equity, and Homelessness Austin AI Housing Analysis Austin AI Housing Analysis Key aspects Housing affordability model Geographic and demographic analysis of change over time Housing costs + Energy costs + Transportation costs Tool for evaluating and testing equitable regulatory scenarios Housing affordability and COVID-19 (what has changed so far?) Austin AI Housing Analysis Milestones 1. Data Repository ○ Web-based open data portal ○ ○ ○ ○ ○ ○ ○ ○ 2. Longitudinal Housing Affordability Model (Historical Analysis) Change in appraised property values, as well as energy and transportation costs 1990 to 2020, yearly Parcel scale (analysis of types of housing development) Analysis of outcomes of past policy (what has helped?) Model will provide training data for predictive AI 3. Regulatory Scenario Equity Evaluation (Predictive Analysis) AI-assisted Projection of past affordability trends into future Estimate effect of various regulation changes on affordability projections based on past outcomes 4. Public workshops to introduce analysis data and method to researchers COA-UT Good Systems Projects Project Exploration & Team Formation Workshop Overview • Event held February 2020 • Cocreated by UT and COA staff • Design thinking methodology • 45 COA participants / 19 departments • 36 UT participants / 15 departments Outcomes • Networking • Cross-pollination of ideas​ • Better understanding of COA priority challenges​ & UT expertise and interests​ • Initial project idea and team formation​ + 1 month to formalize and submit a proposal Project Teambuilding and Strategy Workshop Overview • Event held November 2020 • Co-organized by UT and COA • Remote strategic cocreation workshop Outcomes • Project alignment with Smart Cities • Better understanding of stakeholder needs • Team building • Co-created collaboration plans Next Up: Project exploration and budget planning workshop - February 2021 City of Austin, University of Texas Formalize Research Partnership • 5-year interlocal agreement • Pre-approves $7.5 million for research, consulting and technical assistance from UT faculty and researchers • Pre-negotiated terms and conditions • Removes administrative barriers & streamlines city research "The new ILA facilitates much more rapid collaboration between UT experts and City staff, while collecting tracking information and progress on these collaborations in one location, which should help to ensure closer coordination across projects to improve accountability and to reduce duplicative efforts and costs.” - Jennifer Lyon Gardner UT Deputy VP for Research “With the university as a partner, the City has a powerful resource for addressing our most challenging civic issues. The ILA is the right move to set the City up for long- term success in collaborative research and data-driven decision making,” - Mayor Steve Adler Big Picture City of Austin & Good Systems Plans for the Future What are we doing now and where will we be in the future? AI Vision for Austin: A community-wide competency & AI literacy Increased efficiency, better decision making • • Community partnerships to ensure safe, equitable outcomes Mutual Benefit: domain knowledge & knowledge sharing • Interdisciplinary and transdisciplinary teams solve real-world challenges Example future project: UT, CTM and Equity Office collaboration • Research on bias and AI procurement and implementation How the CTTC can help • Ongoing support to continue Good Systems work • Project formation and priority alignment • Budget assistance and recommendations • Staffing & competency needs to support AI long-term • Active participation and advocacy Thank you!