A non-technical introduction to the state of the art in Artificial Intelligence (AI) and Machine Learning (ML), with particular emphasis on their current applications in the fields of Architecture, Landscape, Urbanism, and Real Estate. Participants will acquire the skills to manage a team in a successful machine learning project, without needing the expertise to understand the details of its technical implementation.
MaP+S Group’s Chuck Hoberman along with Justin Werfel (Wyss), and Robert Wood (Wyss) to lead development of robotic strategies able to maintain crewed and un-crewed extra-terrestrial habitats in space and on Mars. This multidisciplinary effort, called the Resilient ExtraTerrestrial Habitats institute (RETHi), is led by a team from Purdue University in partnership with Harvard and additional teams from the University of Connecticut and the University of Texas at San Antonio, and will receive as much as $15 million over a five-year period.
Hold the date for an intriguing day of conversations and presentations. The conference will be held on March 27 at the Radcliff Gym at Harvard – the event is free and open to the public. Breakfast and lunch will be served. Description:
Science and design are often thought to represent opposite mindsets of analysis versus conjecture, of explanatory versus disruptive modes of thinking. Audacious, provocative, and unconventional work emerges when we cross the traditional disciplinary divide between science and design. The March 2020 conference assembles thought leaders from both areas—researchers and practitioners who interrogate the spaces in between through their work. The sessions, which are thematically aligned with the human senses, present a multitude of approaches, methods, and techniques that transcend the disciplinary focus. Conceived as an open forum for exploration and discussion, the conference is co-hosted by faculty from the Graduate School of Design and the John A. Paulson School of Engineering and Applied Science at Harvard University. The projects and explorations featured range from fashion to material science, from therapeutic devices to building science, and from cooking to cognitive science.
We welcome you to join us for the 2020 DDes Conference held on February 26th at the GSD, Stubbins (Room 112) from 6:30 pm to 8:30 pm.
The conference, titled, “Architecture as a ‘Learning’ Living Machine” will address new emerging concepts in the field of data-driven, adaptive and robotic architecture such as ML, Robotic Buildings, Smart Homes in a form of a symposium.
We will be presenting this year’s ceramic installation, our tallest ever, at the Cevisama trade fair in Valencia from 3 – 7 February! Hope to see you there! We thank the Cevisama Trade Fair Organization and ASCER Tile of Spain for their support.
Mark your calendar for our first Design – Science Conference at Harvard, showcasing intriguing and compelling work at the intersection of design and science at Harvard and beyond. Curated by Allen Sayegh, Joanna Aizenberg and Martin Bechthold, the event is held at the Radcliff Institute for Advanced Studies, free and open to the public. more
We just published a study on the use of Google Street View, combined with U.S. Census data, to create meaningful datasets that can be used for accurate LCA of buildings. Check out the paper and LinkedIn post.
We hope to see many of you at ACADIA 2019: Ubiquity and Autonomy in Austin, Texas on Oct 24-26. We will be presenting a paper on Robot Ex Machina, a computational framework on concurrent human-robot interaction and one on Janus Printing: Co-extrusion based Multi-material 3D printing of Ceramics! Publications will be linked soon. Link
MaP+S Group would like to welcome Dr. Iryna Kuksa as our visiting scholar for the year. Iryna Kuksa is a Senior Research Fellow at the School of Art and Design, Nottingham Trent University (UK). She received her Doctorate from the School of Theatre, Performance and Cultural Policy Studies, University of Warwick. Currently, Iryna is a Visiting Scholar at The Harvard Graduate School of Design, working on an authored book Understanding Personalization: New Aspects of Design, Consumption and Creation (Elsevier, 2020).
Congratulations to Zach Seibold and Jose Luis García del Castillo y López for both joining the faculty this year! Zach will be teaching in the MDE program, as well as offering ‘Digital Media: Manipulations‘ (co-taught with Hyojin Kwon, a research associate with the Laboratory for Design Technologies). Jose Luis is teaching the MDE tech required ‘Introduction to Computational Design‘ as well as ‘Structural Design I‘ with Martin Bechthold.
Our cold-SNAP project, an innovative evaporative cooling system developed jointly with Prof. Aizenberg’s Adaptive Materials Platform (involving Jack Alvarenga), the Harvard Center for Green Buildings and Cities (involving Jonathan Grinham and Saurabh Mhatre), and MaP+S, was featured in a story issued by the Wyss institute. Cold-SNAP was recently patented.
We are happy to participate, along with our colleagues Andrew Witt and Allen Sayegh, in the Aerial Futures Conference at Harvard and MIT. Short distance autonomous flight is technically possible but significant questions remain on where and how it can add value to society. We are thrilled to be part of the conversation!
MaP+S has filed an International PCT Patent on a deployable mechanism with applications ranging from medical devices, wheels, furniture, tents to deployable shelters. Related publications are forthcoming.
Mhatre, S. A., Bechthold, M., & Sayegh, A. ‘ DEPLOYABLE KIRIFORM FLEXURES. Patent No. WO2019018546, 2019
Prof. Bechthold is lecturing on ‘Multi-Scalar Design: Material Matters’ at Georgia Tech on January 15 at noon in the Flex Space, Caddell Building, 280 Ferst Drive, Atlanta, GA 30332-0680. The talk focuses on the challenges and opportunities provided by the expanded range of scales that design and science interact in.
MaP+S is founding partner of the Laboratory for Design Technologies (LDT), a new research organization at the GSD. LDT is the umbrella organization for all technology related GSD design labs such as REAL (Allen Sayegh), the Geometry Lab (Andrew Witt) and the City Form Lab (Andres Sevtsuk). LDT also integrates work by our colleagues currently not associated with a particular design lab – for example Prof. Sawako Kaijima. LDT is currently supported by several industry advisors that include Perkins & Will, Hickok Cole Architects, and Advanced Multitech Corp. We expect to be announcing other industry partners shortly…
Chuck Hoberman’s course review is testing small models and large prototypes of pneumatically actuated origami systems that could potentially assemble into building-scale kinetic structures. Hoberman’s work is expanding the known approaches to kinetic building systems, with potential applications in other fields such as consumer products.
We hope to see many of you at ACADIA 2018 in Mexico City. We will be presenting a paper on Ferrofluidic Casting and two on the 3D printing of ceramics!
• Responsive Spatial Print: Clay 3D printing of spatial lattices using real-time model recalibration. Hyeonji Claire Im; Sulaiman Alothman; Jose Luis García del Castillo (Thur 9.30am)
• Fluid Equilibrium: Material computation in ferrofluidic castings. Zach Seibold; Jonathan Grinham; Olga Geletina; Onye Ahanotu; Allen Sayegh; James Weaver; Martin Bechthold (Fri 4.45 pm)
• Ceramic Morphologies: Precision and control in paste-based additive manufacturing. Zach Seibold; Kevin Hinz; José Luis García del Castillo y López; Nono Martínez Alonso; Saurabh Mhatre; Martin Bechthold (Sat 3 pm)
Doctoral candidate Jose Luis García del Castillo will be co-teaching the “Talk to a Wall” workshop at the forthcoming ACADIA 2018 conference in Mexico City.
This workshop will be a collaborative design exercise which seeks to explore new pluralities on the intersubjectivities of human and machine within the context of architectural design methods. Participants will train a neural network (LSTM) by procedurally sketching architectural ideas, they will then select words and images that associate to those sketches. The network will learn from the participant drawings and do searches in a higher dimensional space to understand the associations between the semantics and their drawings. The model should then be able to draw in a similar style to that of the participant when triggered by a concept word or by an initial set of parameters. The model predictions will then be translated into robotic motion parameters to drive the kinematics of a 6-axis industrial robotic arm. Participants will gain skills in procedurally sketching methods, machine learning and robotic fabrication within a framework of bidirectional communication between both.
The workshop will be led by Alicia Nahmad (AADRL), Vishu Bhooshan (Zaha Hadid Architects CODE), Cristobal Valenzuela (NYU, RunwayML) and Jose Luis, and will be held at the School of Architecture at the Universidad Nacional Autónoma de México (UNAM) October 15-17th 2018.