Tag

research computing

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 MOSAIC 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)

ARC-TS Town Hall on Next Generation HPC Cluster

By |

The University of Michigan is beginning the process of building our next generation HPC platform, “Big House.”  Flux, the shared HPC cluster, has reached the end of its useful life. Flux has served us well for more than five years, but as we move forward with replacement, we want to make sure we’re meeting the needs of the research community.

ARC-TS will be holding a series of town halls to take input from faculty and researchers on the next HPC platform to be built by the University.  These town halls are open to anyone.

Your input will help to ensure that U-M is on course for providing HPC, so we hope you will make time to attend one of these sessions. If you cannot attend, please email hpc-support@umich.edu with any input you want to share.

ARC-TS Town Hall on Next Generation HPC Cluster

By |

The University of Michigan is beginning the process of building our next generation HPC platform, “Big House.”  Flux, the shared HPC cluster, has reached the end of its useful life. Flux has served us well for more than five years, but as we move forward with replacement, we want to make sure we’re meeting the needs of the research community.

ARC-TS will be holding a series of town halls to take input from faculty and researchers on the next HPC platform to be built by the University.  These town halls are open to anyone.

Your input will help to ensure that U-M is on course for providing HPC, so we hope you will make time to attend one of these sessions. If you cannot attend, please email hpc-support@umich.edu with any input you want to share.

ARC-TS Town Hall on Next Generation HPC Cluster

By |

The University of Michigan is beginning the process of building our next generation HPC platform, “Big House.”  Flux, the shared HPC cluster, has reached the end of its useful life. Flux has served us well for more than five years, but as we move forward with replacement, we want to make sure we’re meeting the needs of the research community.

ARC-TS will be holding a series of town halls to take input from faculty and researchers on the next HPC platform to be built by the University.  These town halls are open to anyone.

Your input will help to ensure that U-M is on course for providing HPC, so we hope you will make time to attend one of these sessions. If you cannot attend, please email hpc-support@umich.edu with any input you want to share.

ARC-TS Town Hall on Next Generation HPC Cluster

By |

The University of Michigan is beginning the process of building our next generation HPC platform, “Big House.”  Flux, the shared HPC cluster, has reached the end of its useful life. Flux has served us well for more than five years, but as we move forward with replacement, we want to make sure we’re meeting the needs of the research community.

ARC-TS will be holding a series of town halls to take input from faculty and researchers on the next HPC platform to be built by the University.  These town halls are open to anyone.

Your input will help to ensure that U-M is on course for providing HPC, so we hope you will make time to attend one of these sessions. If you cannot attend, please email hpc-support@umich.edu with any input you want to share.

U-M, Yottabyte partner to accelerate data-intensive research

By | General Interest, News

CONTACT: Dan Meisler, ARC Communications Manager, 734-764-7414, dmeisler@umich.edu

A strategic partnership between the University of Michigan and software company Yottabyte promises to unleash a new wave of data-intensive research by providing a flexible computing cloud for complex computational analyses of sensitive and restricted data.

The Yottabyte Research Cloud will provide scientists high performance, secure and flexible computing environments that enable the analysis of sensitive data sets restricted by federal privacy laws, proprietary access agreements, or confidentiality requirements. Previously, the complexity of building secure and project-specific IT platforms often made the computational analysis of sensitive data prohibitively costly and time consuming.

The system is built on $5.5 million worth of hardware and software donated to the University by Yottabyte; U-M will provide $2 million to support delivery of services to researchers and general operations.

Brahmajee Nallamothu, professor of internal medicine, tested a pilot installation of the Yottabyte Research Cloud at the U-M Institute of Healthcare Policy and Innovation for his research on such topics as predictors of opioid use after surgery and the costs and uses of cancer screenings under the Affordable Care Act.

“We recently moved a healthcare claims database, which is multiple terabytes in size and requires a great deal of memory and fast storage to process, onto the pilot platform,” Nallamothu said. “The platform allows us to immediately increase or decrease computing resources to meet demand while permitting multiple users to access the data safely and remotely. Our previous setup relied on network storage and self-managed hardware, which was extremely inefficient compared to what we can do now.”

“The Yottabyte Research Cloud will improve research productivity by reducing the cost and time required to create the individualized, secure computing platforms that are increasingly necessary to support scientific discovery in the age of Big Data,” said Eric Michielssen, associate vice president for advanced research computing at U-M.

“With the Yottabyte Research Cloud, researchers will be able to ask more questions, faster, of the ever-expanding and massive sets of data collected for their work,” said Yottabyte CEO Paul E. Hodges, III. “We are very pleased to be a part of the diverse and challenging research environment at U-M. This partnership is a great opportunity to develop and refine computing tools that will increase the productivity of U-M’s world class researchers.”

Many U-M scientists are working on a variety of research projects that could benefit from use of the Yottabyte Research Cloud:

  • Healthcare research, for example in precision medicine, often requires working with sensitive patient information and large volumes of diverse data types. This research can yield results that positively impact patients’ lives, but often involves the analysis of millions of clinical observations that can include genomic, hospital, outpatient, pharmaceutical, laboratory and cost data. This requires a secure high performance computing ecosystem coupled to massive amounts of multi-tiered storage.
  • In the social sciences, U-M research requires secure, remote access to sensitive research data about substance abuse, mental health, and other topics.
  • Transportation researchers who mine large and sensitive datasets — for example, a 24 Terabyte dataset that includes videos of drivers’ faces and GPS traces of their journeys — also stand to benefit from the security features and computing power.
  • In learning analytics, studies of the persistence of teacher effects on student learning could benefit from the enclaves to store and analyze data that includes observational measures scored from classroom videos, and elementary and middle school students’ scores on standardized tests.
  • Researchers in brain science will be able to use the Yottabyte Research Cloud to investigate a wide range of topics including  the effects of aging on brain function and structure and how we focus our attention in the presence of distraction.

The Yottabyte Research Cloud is U-M’s first foray into software-defined infrastructure for research, allowing on-the-fly personalized configuration of any-scale computing resources, which promises to change the way traditional IT infrastructure systems are deployed across the research community.  

More about Yottabyte:  www.yottabyte.com.

More about Yottabyte Research Cloud: arc-ts.umich.edu/yrc

Questions: dmeisler@umich.edu

Video, slides available: “Advanced Research Computing at Michigan, An Overview,” Brock Palen, ARC-TS

By | General Interest, News

Video (http://myumi.ch/aAG7x) and slides (http://myumi.ch/aV7kz) are now available from Advanced Research Computing – Technology Services (ARC-TS) Associate Director Brock Palen’s presentation “Advanced Research Computing at Michigan, An Overview.”

Palen gave the talk on June 27, 2016, outlining the resources and services available from ARC-TS as well as from off-campus resource providers.

Advanced Research Computing at Michigan — An Overview

By |

Brock Palen, Associate Director of Advanced Research Computing – Technology Services, will provide an overview of the resources available to researchers engaged in computationally intensive science on the University of Michigan campus.

The talk is open to researchers from any department at U-M.

The session will address:

  • high performance computing services
  • data science services such as Hadoop and Spark
  • research storage
  • cloud services
  • database hosting
  • networking services
  • grant consultation and collaboration
  • access to off-campus resources.

There will be time for questions and answers after the presentation.

RSVP requested.

A live video stream of the event will be available: