Projects

September 17, 2023, Filed Under: Digital twin, News & Events, Projects

A Digital Twin for Austin

Announcing the Austin Digital Twin: A Revolutionary Leap in Urban Management and Public Safety

We are thrilled to announce the launch of the Austin Digital Twin, a groundbreaking initiative aimed at transforming how cities are managed and safeguarded. Developed through a collaborative effort between the University of Texas Urban Information Lab, the City of Austin, and the TeXuS Lab, the Austin Digital Twin is an interactive, real-time 3D map that offers unparalleled insights into the dynamic urban landscape of Austin, Texas.

Key Features

– Real-Time Monitoring: Receive immediate updates on critical urban elements such as fires, police activity, and live traffic.
– Informed Decision-Making: The fusion of real-time data streams empowers both residents and authorities with an unprecedented level of situational awareness.
– Enhanced Public Safety: The digital twin facilitates proactive public safety measures by enabling swift responses to emergency situations.

We invite you to explore this cutting-edge technology by visiting AustinDigitalTwin.com.

Previous Engagements
Our team from the Urban Information Lab had the privilege to present the Austin Digital Twin project at the recently concluded Smart Cities and Generative AI Symposium. The symposium was a gathering of interdisciplinary professionals who examined the impact of generative artificial intelligence on urban environments. We are pleased to share that the project received significant attention and fostered insightful discussions on the role of technology in shaping smarter, safer cities.

Further Research and Development
Interested in our other digital twin initiatives? You’re invited to explore our range of projects on GitHub, where you can find repositories such as FireIncidentData, OpenCityAI, and more. These projects continue our commitment to leveraging advanced technologies for the betterment of urban living.

We look forward to your participation and support as we continue to develop technologies that contribute to the creation of resilient, sustainable, and intelligent urban systems.

For more information, please visit AustinDigitalTwin.com and our GitHub page.

Share this:

May 17, 2023, Filed Under: Artificial Intelligence, Projects

ComputeGPT

ComputeGPT is a computational chat model that accurately answers math and science problems. It outperforms GPT-4 with Internet (Bing), GPT-3.5 (ChatGPT), GPT-3, and Wolfram Alpha NLP, scoring 98% across a suite of numerical problems that Bing AI could only answer at 64% and Wolfram Alpha NLP at 56%.

One of the focuses for this project was on accuracy, as current AI solutions have been shown to have inaccurate or misleading answers, known as “hallucinations”. Our model runs verifiably correct code on-demand to ensure that the answer it gives is completely accurate. This project fixes the “probabilistic accuracy” in language models created by LLM’s inherent architecture: a focus on generating the next word. By instead generating code and running it, ComputeGPT demonstrates state-of-the-art performance in homework, textbook, and all kinds of numerical problems, including mathematics, physics and medicine. ComputeGPT is a math chat model for computing any numerical answer.

See our paper here: https://arxiv.org/abs/2305.06223.
See our GitHub here: https://github.com/urbaninfolab/ComputeGPT.

Unlike other LLMs, ComputeGPT is available at no cost and logs no information about its users. Try it out now at https://computegpt.org.

Team Members:
Junfeng Jiao (Department of Architecture), Ryan Hardesty Lewis (Department of Mathematics)

(Click Here for More Project Information)

Math chat model MathGPT ComputeGPT — ComputeGPT is a state-of-the-art math chat model.

Share this:

April 21, 2023, Filed Under: Artificial Intelligence, Projects

OpenCityAI

OpenCityAI is a AI chat model trained on a massive amount of public city information, making it able to generate answers for a variety of questions that people may have about their cities. It provides accurate responses with citations in a simple easy-to-use interface. One of the focuses for this project was on accuracy, as current AI solutions have been shown to have inaccurate or misleading answers. Our model has been validated at being exceptionally more effective at answering city-based data questions as compared to other AIs. This project bridges the gap of being able to find specific information about a city, and allows anyone to access accurate information about any city immediately.

Team Members:
Junfeng Jiao (Principal Investigator, Architecture), Weijia Xu (Texas Advanced Computing Center), Kijin Seong (Urban and Regional Sciences), Sean Henry Lewis (Department of Mathematics)

(Click Here for More Project Information)

Share this:

January 1, 2023, Filed Under: Digital twin, Projects

Smart City Data Integration

Open Data portals and such for cities are increasingly popular for local goverment to publish and communicate data to the public. While the dedicated interface makes it easy for data owners to upload data, it lacks comprehensive means to connect relavant data for a particular topic of interest. As datasets are maintained by individuals, they often lack of conherent schema and provide different levels of detail in the description. As a result, the more data an open data portal has, the harder it gets for interested users to identify and use the data. In this project, we aim to start bridging the gaps between diverse data sets and their usability to the public. Our aims include creating an interface to connect and search an open data portal over mulitple cities, identifying data related to specific topic of interests, and generate both spatial and knowledge maps to show how the data is connected and distributed.

Interactive Data Model Example

Team Members:
Junfeng Jiao (Principal Investigator, Architecture), Weijia Xu (Texas Advanced Computing Center), Kijin Seong (Urban and Regional Planning), Yingying Xu (Department of Computer Science), Sean Henry Lewis (Department of Mathematics)

(Click Here for More Project Information)

Share this:

December 19, 2022, Filed Under: Projects

National Housing Data Portal

Open Data portal containing housing data from realtor.com and additional city information. Users can choose to view aggregated housing data from the realtor.com residential listings database or market analysis metrics from realtor.com market hotness index. We enrich this data with US Census demographic data and geometric shp files. Users can access, filter, aggregate, visualize, and download this data from any city of interest in the nation.

Team Members:
Prachi Ingle (Dept. of Computer Science), Junfeng Jiao (Architecture), Arya Farahi (Dept. of Statistics and Data Science)

Share this:

December 19, 2022, Filed Under: Projects

Community Hub for Smart Mobility (CHSM)

The goal of this project is to develop, implement, and evaluate a Community Hub for Smart Mobility (CHSM) to solve the job/housing mismatch in US cities. The research team will focus on Georgian Acres, a historically under-resourced neighborhood in northeast Austin, Texas and work with multiple civic partners to address the unique transportation needs of the neighborhood.

Team Members:
Junfeng Jiao (Principal Investigator, Architecture), Jason JonMichael (Smart Mobility), Kenneth Fleischmann (School of Information), Sherri Greenberg (School of Public Affairs), Devrim Ikizler (Computational Economics), Catherine Cubbin (School of Social Work), Ming Zhang (School of Architecture), Hannah Barron (Smart Mobility), Alex Pyson (Smart Mobility), Chris Haskins (Smart Mobility), Michael Salazar (Smart Mobility), Amin Azimian (Civil Engineering), Shunhua Bai (Community and Regional Planning), Charles Bell (Engineering), Rand Makarem (Urban Planning), Marsha Phillips (Community and Regional Planning), Abby Johnson (Economics), Yefu Chen (Community and Regional Planning)

(Click Here for More Project Information)

Share this:

December 19, 2022, Filed Under: Deserts, Projects

Austin AI Housing Analysis

A lack of affordable housing is a major problem in US cities from the Bay Area to Boston. Austin is no exception. In 2015, the Austin-Round Rock metropolitan area was named one of the most economically segregated areas in the country. With more people moving to the region – roughly 3,000 people per month – the disparity will worsen. This research project will develop a values-driven AI system that evaluates historical housing development and helps policymakers shape equitable, inclusive and sustainable plans and regulations. Using deep learning technology and an open data repository with demographic, development, transportation, and energy consumption data, this system will link and study 50 years of Austin housing development figures in an effort to help tackle the future of local housing. More specifically, the project will look at the spatial and demographic relationships between housing, affordability, mobility and energy use over the past 50 years and how likely those patterns are to persist depending on what regulations are put in place. The goal is to build a predictive AI system that will test different regulatory scenarios and inform residential development policies in Austin. The techniques and findings developed from this study can also be applied to many similar cities nationally and internationally. Additionally, the data repository created here will be used to study how environmental and socioeconomic factors have affected the spread of COVID-19 in the city of Austin.

Team Members:
Junfeng Jiao (Principal Investigator, Architecture), Weijia Xu (Texas Advanced Computing Center), Michelle Addington (Architecture), Ming Zhang (Architecture), Jake Wegmann (Architecture), Katie Pierce Meyer (University of Texas Libraries), Hao Zhu (Electrical and Computer Engineering), Amin Azimian (Architecture), Josh Conrad (Architecture), Cara Bertron (City of Austin Planning and Zoning Department), Matt Dugan (City of Austin Planning and Zoning Department), John Clary (City of Austin Transportation Department), Molly Emerick (Austin Energy), Matt Hollon (City of Austin Watershed Protection Department) and Jacquie Hrncir (City of Austin Communications & Technology Management Department)

Share this:

December 19, 2022, Filed Under: Artificial Intelligence, Projects

AI Image Generation for Architecture Design

This project allows users to create unique images depicting architecture and design to assist with the idea creation process. Architects often need to create novel, never-before-seen designs for buildings, landscapes, and interiors. With so many existing buildings in the world, it can be difficult to come up with truly unique ideas. The Artificial Intelligence (AI) revolution presents architects with an entire set of new tools that can be used to complement the ideation phase of design. These tools previously have been available only to individuals with a strong software development background. However, recent advances have made these tools accessible to everyone, requiring no technical knowledge or programming skills.

Team Members:
Junfeng Jiao (Principal Investigator, Architecture), Jonathan Mallott (Architecture)

(Click Here for More Project Information)

Share this:

November 3, 2022, Filed Under: Digital twin, Projects

Active Fire Incident Map

This project provides active fire incidents of every major city on an interactive map. The tracker is updated in almost real-time from all over the nation. This live tracker updates with the latest smoke runoff from each fire, keeping you well informed and out of danger, as well as keeps track of the known fire risk of every tract in all tracked cities. It has currently tracked over 17,000 fires and continues to track tens to hundreds of new active fire incidents every day.

Team Members:
Junfeng Jiao (Principal Investigator, Architecture), Weijia Xu (Texas Advanced Computing Center), Dev Niyogi (Geological Sciences), Kijin Seong (Urban and Regional Sciences), Ryan Hardesty Lewis (Department of Mathematics)

(Click Here for More Project Information)

Share this:

September 3, 2022, Filed Under: Artificial Intelligence, Projects

Convergent, Responsible, and Ethical AI Training Experience (CREATE Roboticists)

CREATE Roboticists will train future roboticists who: (i) understand the ethical implications of service robots and can develop new theories, methods, and techniques to satisfy ethical requirements; (ii) design human-centered ethical service robots that respect human autonomy and ethical values; and (iii) develop robotics policy informed by cutting edge convergent research.

Team Members:
Junfeng Jiao (Principal Investigator, Architecture), Luis Sentis (Electrical Engineering), Joydeep Biswas (Computer Science), Min Kyung Lee (School of Information), Justin Hart (Department of Computer Science), Kenneth Fleischmann (School of Information), Peter Stone (Department of Computer Science), Tina Peterson (Department of Computer Science), Ann Martin (Globe Project)

(Click Here for More Project Information)

Share this: