Prof. Neil Gershenfeld
Director
http://cba.mit.edu/~neilg
NSF CCR-0122419
Proposal
Cynthia Breazeal (MAS)
Bill Butera (MAS)
Isaac Chuang (MAS, Physics)
Drew Endy (Bio.)
Neil Gershenfeld (MAS)
Kim Hamad-Schifferli (Mech. E.)
Joseph Jacobson (MAS)
Tom Knight (CSAIL)
Seth Lloyd (Mech. E.)
Scott Manalis (MAS, Bio. E.)
Bakhtiar Mikhak (MAS)
Joe Paradiso (MAS)
Pablo Parrilo (EECS)
Sandy Pentland (MAS)
Mitchel Resnick (MAS)
Rahul Sarpeshkar (EECS, RLE)
Larry Sass (Arch.)
Sebastian Seung (BCS, Physics)
Peter Shor (Math)
Alex Slocum (Mech. E.)
Karen Sollins (EECS, LCS)
Timothy M. Swager (Chem.)
Shuguang Zhang (Bio. E.)
CBA People
Sherry Lassiter
Susan Murphy-Bottari
John Difrancesco
Mike Houlihan
Ruzena Bajcsy
Charles Bennett
Barrie Gilbert
Alan Huang
Nathan Myhrvold
Greg Papadopoulos
John Doyle
Year 0
systems
substrates
foundations
Year 1
RF biology
personal fabrication
quantum computing
analog logic
silicon biology
nanogate
sensate surfaces
adaptive robotics
shape grammar
paintable computing
Year 2
RF biology
personal fabrication
quantum computing
it from bit: how can functional description be embodied in physical form?
analog logic
silicon biology
nanogate
bit from it: how can functional description be abstracted from physical form?
sensate surfaces
adaptive robotics
shape grammar
paintable computing
Year 3
it from bit: how can functional description be embodied in physical form?
building with logic
bit from it: how can functional description be abstracted from physical form?
programming with math
Logical Assembly
(Saul Griffith)
Thresholds
1940s: Communications
(Shannon)
errors
1950s: Computation
(Winograd, von Neumann)
2000s: Fabrication
noise
Analog Logic
(Ben Vigoda, Andi Loeliger, ...)
xj
xi
fn
fm
Programming Distributed Systems
problem
algorithm
program
executable
protocol
messages
dynamics
Graphical Message-Passing
problem
algorithm
program
executable
protocol
messages
dynamics
KLM NLL
Internet 0 (I0)
(Raffi Krikorian, Danny Cohen, Doug Johnson)
IR
RF
powerline
multidrop
RFID
bar codes
mag stripe
telephone
telegraph
IRDA
Bluetooth
Homeplug
RS-485
EPC
UPC
ANSI/ISO
V.92
Morse Code
3x108 m/s / 100 m = 3x106 s-1
•
•
•
•
•
•
•
IP to leaf nodes
peers don’t need server
physical identity
compiled standards
open standards
big bits
end to end modulation
interdevice
internetworking
•
•
•
•
•
•
•
•
•
MAS.862: The Physics of Information Technology
MAS.863: How To Make (Almost) Anything
MAS.864: The Nature of Mathematical Modeling
MAS.961: How To Make Something That Makes (almost) Anything
6.151: Semiconductor Devices Project Laboratory
6.971: Engineering Simple Biological Systems
7.86, BE.481, MAS.866: Fundamental limits of biological
measurement
8.371J, MAS.865J: Quantum Information Science
BE.442: Molecular Structure of Biological Materials
CBA
Courses
Maguire
Vigoda
Cambridge Series on Information and the Natural Sciences
Graduate Study in
Design and the Natural Sciences
Design and the Natural Sciences is a graduate academic program asking how the resources of natural
systems can be used to embody functional designs in physical forms, and conversely how functional
descriptions can be abstracted from physical forms. It provides training in the interdisciplinary research
areas associated with MIT's Center for Bits and Atoms (CBA), bringing together faculty from across campus
in departments including Physics, Chemistry, Biology, Mathematics, Computer Science, and Electrical and
Mechanical Engineering, all working at the interface between logical and physical representations of
information. DNS is part of the Media Arts and Sciences (MAS) program, which provides a broader context
for studying the social as well as intellectual impact of emerging technologies on human expression.
DNS teaches design practice in science, rather than scientific practice in design. Herbert Simon first
articulated the goal of a "science of design", in The Sciences of the Artificial. This program sought to create
desired artificial systems rather then describe existing natural ones, and was realized in the development of
CAD and machine optimization, Artificial Intelligence and Artificial Life, and ultimately virtual digital worlds.
The Sciences of the Artificial was itself a response to the growing dominance of physical science in
engineering. The success of science in World War II, including the connection between particle physics and
nuclear weapons, and between microwave spectroscopies and radar, was followed by the growth of
engineering as a scientific rather than empirical discipline. A scientific approach to design was seen as
being needed to counter the rise of experimental studies in new areas such as condensed matter physics,
which emphasized observation over problem-solving skills.
This split between description and prescription can be traced still further back, to the emergence of the
modern notion of literacy in the Renaissance as a mastery of the available means of expression. This
comprised the language and rhetoric of the Trivium, and the natural science of the Quadrivium; practical
concerns of making things were relegated to the "illiberal arts" as a commerical concern.
DNS seeks to correct this accumulated historical division between the artificial and natural. Abstractions
that isolate the process of design from underlying physical degrees of freedom are increasingly
unsustainable, driven by the demands of fundamental physical scaling limits as well as:
Fabrication
m
mm
μm
nm
in
out
How To Make
(almost) Anything
Boston
Fab Labs
space (m), time (s)
fabrication, instrumentation divides
Ghana, Costa Rica, ...
India
Norway
3D scan/mill
analytical
instrumentation
electronics
PCB,
electromagnetics