Category

Educational

Using GenAI to design floor plans and buildings

By | Data, Data sets, Educational, Feature, General Interest, HPC, News, Research, Systems and Services

There is a lot to consider when designing places where humans live and work. How will the space be used? Who’s using the space? What are budget considerations? It is painstaking and time consuming to develop all of those details into something usable. 

What if Generative AI (GenAI) could help? We already know that it can be used to create text, music, and images. Did you know that it can also create building designs and floor plans? 

Dr. Matias del Campo, associate professor of architecture in the Taubman College for Architecture and Urban Planning, has been working to make architectural generative models more robust. He aims to expand on the patterns, structures, and features from the available input data to create architectural works. Himself a registered architect, designer, and educator, del Campo conducts research on advanced design methods in architecture using artificial intelligence techniques.

He leverages something called neural networks for two projects: 

  • Common House: A project that focuses on floor plan analysis and generation.
  • Model Mine: A large-scale, 3D model housing database for architecture design using Graph Convolutional Neural Networks and 3D Generative Adversarial Networks.

This is an example from the annotated data created from the Common House research project. The main obstacle that has emerged in creating more real-life plans is the lack of databases that are tailored for these architecture applications. The Common House project aims at creating a large-scale dataset for plans with semantic information. Precisely, our data creation pipeline consists of annotating different components of a floor plan, for e.g., Dining Room, Kitchen, Bed Room, etc.

 

Four quadrants showing 9 models each of chairs, laptops, benches, and airplanes

A large scale 3D model housing database for Architecture design using Graph Convolutional Neural Networks and 3D Generative Adversarial Networks.

What exactly are neural networks? The name itself takes inspiration from the human brain and the way that biological neurons signal to one another. In the GenAI world, neural networks are a subset of machine learning and are at the heart of deep learning algorithms. This image of AI hierarchy may be helpful in understanding how they are connected.

Dr. del Campo’s research uses GenAI for every step of the design process including 2D models for things like floors and exteriors, and 3D models for shapes of the rooms, buildings, and volume of the room. The analysis informs design decisions. 

DEI considerations

del Campo notes that there are some DEI implications for the tools he’s developing. “One of the observations that brought us to develop the ‘Common House’ (Plangenerator) project is that the existing apartment and house plan datasets are heavily biased towards European and U.S. housing. They do not contain plans from other regions of the world; thus, most cultures are underrepresented.” 

To counterbalance that, del Campo and his team made a global data collection effort, collecting plans and having them labeled by local architects and architecture students. “This not only ensured a more diverse dataset but also increased the quality of the semantic information in the dataset.”

How technology supports del Campo’s work

A number of services from Information Technology & Services are used in these projects, including: Google at U-M collaboration tools, GenAI, Amazon Web Services at U-M (AWS), and GitHub at U-M

Also from ITS, the Advanced Research Computing (ARC) team provides support to del Campo’s work. 

“We requested allocations from the U-M Research Computing Package for high-performance computing (HPC) services in order to train two models. One focuses on the ‘Common House’ plan generator, and the other focuses on the ‘Model Mine’ dataset to create 3D models based,” said del Campo. 

Additionally, they used HPC allocations from the UMRPC in the creation of a large-scale artwork called MOSAIK which consists of over 20,000 AI-generated images, organized in a color gradient. 

A large scale 3D model housing database for Architecture design using Graph Convolutional Neural Networks and 3D Generative Adversarial Networks.

“We used HPC to run the algorithm that organized the images. Due to the necessary high resolution of the image, this was only possible using HPC.”

“Dr. del Campo’s work is really novel, and it is different from the type of research that is usually processed on Great Lakes. I am impressed by the creative ways Dr. del Campo is applying ITS resources in a way that we did not think was possible,” said Brock Palen, director of the ITS Advanced Research Computing. 

Related: Learn about The Architecture + Artificial Intelligence Laboratory (AR2IL)

3-2-1…blast off! COE students use ARC-TS HPC clusters for rocket design

By | Educational, General Interest, Great Lakes, Happenings, HPC, News
MASA team photo

The MASA team has been working with the ARC-TS and the Great Lakes High-Performance Computing Clusters to rapidly iterate simulations. What previously took six hours on another cluster, takes 15 minutes on Great Lakes. (Image courtesy of MASA)

This article was written by Taylor Gribble, the ARC-TS summer 2020 intern. 

The Michigan Aeronautical Science Association (MASA) is a student-run engineering team at U-M that has been designing, building, and launching rockets since its inception in 2003. Since late 2017, MASA has focused on developing liquid-bipropellant rockets—which are rockets that react to a liquid fuel with a liquid oxidizer to produce thrust—in an effort to remain at the forefront of collegiate rocketry. The team is made up of roughly 70 active members including both undergraduate and graduate students who participate year-round.

Since 2018, MASA has been working on the Tangerine Space Machine (TSM) rocket which aims to be the first student-built liquid-bipropellant rocket to ever be launched to space. When completed, the rocket’s all-metal airframe will stand over 25 feet tall. The TSM will reach an altitude of 400,000 feet and will fly to space at over five times the speed of sound.

MASA is building this rocket as part of the Base 11 Space Challenge which was organized by the Base 11 Organization to encourage high school and college students to get involved in STEM fields. The competition has a prize of $1 million, to be awarded to the first team to successfully reach space. MASA is currently leading the competition, having won Phase 1 of the challenge in 2019 with the most promising preliminary rocket design.

Since the start of the TSM project, MASA has made great strides towards achieving its goals. The team has built and tested many parts of the complete system, including custom tanks, electronics, and ground support equipment. In 2020, the experimental rocket engine designed by MASA for the rocket broke the student thrust record when it was tested, validating the work that the team had put into the test.

The team’s rapid progress was made possible in-part by the extensive and lightning-quick simulations using the ARC-TS Great Lakes High-Performance Computing Cluster.

The student engineers are Edward Tang, Tommy Woodbury, and Theo Rulko, and they have been part of MASA for over two years.

Tang is MASA’s aerodynamics and recovery lead and a junior studying aerospace engineering with a minor in computer science. His team is working to develop advanced in-house flight simulation software to predict how the rocket will behave during its trip to space.

“Working on the Great Lakes HPC Cluster allows us to do simulations that we can’t do anywhere else. The simulations are complicated and can be difficult to run. We have to check it, and do it again; over and over and over,” said Tang. “The previous computer we used would take as long as six hours to render simulations. It took 15 minutes on Great Lakes.”

A computer simulation of Liquid Oxygen Dome Coupled Thermal-Structural

This image shows a Liquid Oxygen Dome Coupled Thermal-Structural simulation that was created on the ARC-TS Great Lakes HPC Cluster. (Image courtesy of MASA)

Rulko, the team’s president, is a junior studying aerospace engineering with a minor in materials science and engineering.

Just like Tang, Rulko has experience using the Great Lakes cluster. “Almost every MASA subteam has benefited from access to Great Lakes. For example, the Structures team has used it for Finite Element Analysis simulations of complicated assemblies to make them as lightweight and strong as possible, and the Propulsion team has used it for Computational Fluid Dynamics simulations to optimize the flow of propellants through the engine injector. These are both key parts of what it takes to design a rocket to go to space which we just wouldn’t be able to realistically do without access to the tools provided by ARC-TS.”

Rulko’s goals for the team include focusing on developing as much hardware/software as possible in-house so that members can control and understand the entire process. He believes MASA is about more than just building rockets; his goal for the team is to teach members about custom design and fabrication and to make sure that they learn the problem-solving skills they need to tackle real-world engineering challenges. “We want to achieve what no other student team has.”

MASA has recently faced unforeseen challenges due to the COVID-19 pandemic that threaten to hurt not only the team’s timeline but also to derail the team’s cohesiveness. “Beaucase of the pandemic, the team is dispersed literally all over the world. Working with ARC-TS has benefitted the entire team. The system has helped us streamline and optimize our workflow, and has made it easy to connect to Great Lakes, which allows us to rapidly develop and iterate our simulations while working remotely from anywhere,” said Tang. “The platform has been key to allowing us to continue to make progress during these difficult times.”

Tommy Woodbury is a senior studying aerospace engineering. Throughout his time on MASA he has been able to develop many skills. “MASA is what has made my time here at Michigan a really positive experience. Having a group of highly-motivated and supportive individuals has undoubtedly been one of the biggest factors in my success transferring to Michigan.

This image depicts the Liquid Rocket Engine Injector simulation.

This image depicts the Liquid Rocket Engine Injector simulation. (Image courtesy of MASA)

ARC-TS is a division of Information and Technology Services. Great Lakes is available without charge for student teams and organizations who need HPC resources. This program aims to enable students access to high-performance computing to enhance their team’s mission.

Women in HPC launches mentoring program

By | Educational, General Interest, HPC, News

Women in High Performance Computing (WHPC) has launched a year-round mentoring program, providing a framework for women to provide or receive mentorship in high performance computing. Read more about the program at https://womeninhpc.org/2019/03/mentoring-programme-2019/

WHPC was created with the vision to encourage women to participate in the HPC community by providing fellowship, education, and support to women and the organizations that employ them. Through collaboration and networking, WHPC strives to bring together women in HPC and technical computing while encouraging women to engage in outreach activities and improve the visibility of inspirational role models.

The University of Michigan has been recognized as one of the first Chapters in the new Women in High Performance Computing (WHPC) Pilot Program. Read more about U-M’s chapter at https://arc.umich.edu/whpc/

New campus-wide access to MATLAB

By | Educational, General Interest, News

U-M is offering a new, campus-wide license for MATLAB, Simulink, and companion products. All faculty, researchers, and students are eligible to download and install these products, including toolboxes such as:

  • Bioinformatics Toolbox
  • Control System Toolbox
  • Curve Fitting Toolbox
  • Data Acquisition Toolbox
  • Image Processing Toolbox
  • Instrument Control Toolbox
  • Optimization Toolbox
  • Parallel Computing Toolbox
  • Signal Processing Toolbox
  • Simscape
  • Simscape Multibody
  • Simulink Control Design
  • Stateflow
  • Statistics and Machine Learning Toolbox
  • Symbolic Math Toolbox.

Access free, self-paced training to get started in less than 2 hours:  MATLAB Onramp.

Commercial use of MathWorks products is not covered by our TAH license, so if you are using a commercial license, please continue to do so. 

Read more…

Most CSCAR workshops will be free for the U-M community starting in January 2019

By | Educational, General Interest, Happenings, News

Beginning in January of 2019, most of CSCAR’s workshops will be offered free of charge to UM students, faculty, and staff.

CSCAR is able to do this thanks to funding from UM’s Data Science Initiative.  Registration for CSCAR workshops is still required, and seats are limited.

CSCAR requests that participants please cancel their registration if they decide not to attend a workshop for which they have previously registered.

Note that a small number of workshops hosted by CSCAR but taught by non-CSCAR personnel will continue to have a fee, and fees will continue to apply for people who are not UM students, faculty or staff.

U-M approves new graduate certificate in computational neuroscience

By | Educational, General Interest, Happenings, News

The new Graduate Certificate in Computational Neuroscience will help bridge the gap between experimentally focused studies and quantitative modeling and analysis, giving graduate students a chance to broaden their skill sets in the diversifying field of brain science.

“The broad, practical training provided in this certificate program will help prepare both quantitatively focused and lab-based students for the increasingly cross-disciplinary job market in neuroscience,” said Victoria Booth, Professor of Mathematics and Associate Professor of Anesthesiology, who will oversee the program.

To earn the certificate, students will be required to take core computational neuroscience courses and cross-disciplinary courses outside of their home departments; participate in a specialized interdisciplinary journal club; and complete a practicum.

Cross-discplinary courses will depend on a student’s focus: students in experimental neuroscience programs will take quantitative coursework, and students in quantitative science programs such as physics, biophysics, mathematics and engineering will take neuroscience coursework.

The certificate was approved this fall, and will be jointly administered by the Neuroscience Graduate Program (NGP) and the Michigan Institute for Computational Discovery and Engineering (MICDE).

For more information, visit micde.umich.edu/comput-neuro-certificate. Enrollment is not yet open, but information sessions will be scheduled early next year. Please register for the program’s mailing list if you’re interested.

Along with the Graduate Certificate in Computational Neuroscience, U-M offers several other graduate programs aimed at training students in computational and data-intensive science, including:

  • The Graduate Certificate in Computational Discovery and Engineering, which is focused on quantitative and computing techniques that can be applied broadly to all sciences.
  • The Graduate Certificate in Data Science, which specializes in statistical and computational methods required to analyze large data sets.
  • The Ph.D in Scientific Computing, intended for students who will make extensive use of large-scale computation, computational methods, or algorithms for advanced computer architectures in their doctoral studies. This degree is awarded jointly with an existing program, so that a student receives, for example, a Ph.D in Aerospace engineering and Scientific Computing.

 

U-M awarded a Clare Boothe Luce grant for fellowships to support women in STEM

By | Educational, General Interest, Happenings, News

The Clare Boothe Luce Program of the Henry Luce Foundation has awarded a $270,000 grant to the University of Michigan. The funding will support women PhD students through the Michigan Institute for Computational Discovery and Engineering (MICDE). The program aims to encourage women “to enter, study, graduate and teach” in science, and the funding will support women PhD students who make use of computational science in their research.

“We’re very excited to be able to promote women in scientific computing,” said Mariana Carrasco-Teja, manager of the grant and Associate Director of MICDE. “These resources generously provided by the Clare Boothe Luce program will make a huge difference in the careers of women pursuing computational science at U-M.”

For details on applying, and fellowship requirements, see the fellowship page at micde.umich.edu/academic-programs/cbl/.

The fellowships carry a $35,000 annual stipend and tuition, among other benefits. They will be awarded to students applying for PhD programs in fall 2019 in the College of Engineering, or several programs in the College of Literature, Science and the Arts (Applied and Interdisciplinary Mathematics, Applied Physics, Astronomy, Chemistry, Earth & Environmental Sciences, Mathematics, Physics, and Statistics).

The CBL program at U-M is funded by the Clare Boothe Luce Program of the Henry Luce Foundation, with additional support from the Rackham School of Graduate Studies, the College of Engineering, the College of Literature, Sciences and the Arts, and MICDE.

MIDAS announces winners of 2018 poster competition

By | Educational, General Interest, Happenings, Research

The Michigan Institute for Data Science (MIDAS) is pleased to announce the winners of its 2018 poster competition, which is held in conjunction with the MIDAS annual symposium.

The symposium was held on Oct. 9-10, 2018, and the student poster competition had more than 60 entries. The winners, judged by a panel of faculty members, received cash prizes.

Best Overall

Arthur Endsley, “Comparing and timing business cycles and land development trends in U.S. metropolitan housing markets”

Most likely health impact

  • Yehu Chen, Yingsi Jian, Qiucheng Wu, Yichen Yang, “Compressive Big Data Analytics – CBDA: Applications to Biomedical and Health Studies”
  • Jinghui Liu, “An Information Retrieval System with an Iterative Pattern for TREC Precision Medicine”

Most likely transformative science impact

  • Prashant Rajaram, “Bingeability and Ad Tolerance: New Metrics for the Streaming Media Age”
  • Mike Ion, “Learning About the Norms of Teaching Practice: How Can Machine Learning Help Analyze Teachers’ Reactions to Scenarios?”

Most interesting methodological advancement

  • Nina Zhou and Qiucheng Wu, “DataSifter: Statistical Obfuscation of Electronic Health Records and Other Sensitive Datasets”
  • Aniket Deshmukh, “Simple Regret Minimization for Contextual Bandits”

Most likely societal impact

  • Ece Sanci, “Optimization of Food Pantry Locations to Address Food Scarcity in Toledo, OH”
  • Rohail Syed, “Human Perception of Surprise: A User Study”

Most innovative use of data

  • Lan Luo, “Renewable Estimation and Incremental Inference in Generalized Linear Models with Streaming Datasets”
  • Danaja  Maldeniya, “Psychological Response of Communities affected by Natural Disasters in Social Media”

MICDE announces 2018-2019 fellowship recipients

By | Educational, General Interest, Happenings, News

The Michigan Institute for Computational Discovery and Engineering (MICDE) is pleased to announce the 2018-2019 recipients of the MICDE Fellowships for students enrolled in the PhD in Scientific Computing or the Graduate Certificate in Computational Discovery and Engineering. The fellowships, which carry a $4,000 stipend, are meant to augment other sources of funding and are available to students in both programs. See our Fellowship page for more information.

AWARDEES

Zhitong Bai, Mechanical Engineering
Kyle Bushick, Materials Science and Engineering
Geunyeong Byeon, Industrial and Operations Engineering
Sehwan Chung, Civil and Environmental Engineering
Khoi Dang, Chemistry
Sicen Du, Materials Science and Engineering
Joseph Hollowed, Physics
Jia Li, Physics
Sabrina Lynch, Biomedical Engineering
Samar Minallah, Climate and Space Sciences and Engineering
Everardo Olide, Applied Physics
Shaowu Pan, Aerospace Engineering
Alicia Petersen, Climate and Space Sciences and Engineering
Vyas Ramasubramani, Chemical Engineering
Fabricio Vasselai, Political Science
Nathan Vaughn, Applied and Interdisciplinary Mathematics
Blair Winograd, Chemistry
Samuel Young, Chemical Engineering
Kexin Zhang, Chemistry
Bu Zhao, School of Environment and Sustainability