USENIX + MIT Media Lab's having a WS on Embedded Systems

Rohit Khare (
Thu, 28 Jan 1999 12:51:00 -0800

Got this in the mail from USENIX's e-notify list:

March 29-31, 1999, Cambridge, Massachusetts
Co-sponsored by the MIT Media Laboratory
Call for Papers is online. Abstracts due: January 31, 1999
Attendance is limited to 60 selected individuals.

Here's the URL it leads to:

January 31, 1999: Extended abstracts due
February 10, 1999: Acceptance notification
March 15, 1999: Full papers due
March 29, 1999, 12pm: Arrival, Registration (at MIT Media Lab)
March 31, 1999, 12pm: Conclusion

Dan Geer,, CertCo
Mike Hawley,, MIT Media Lab

Program Committee
Mark D. Baushke,, Cisco Systems
Warren Bosch,, Hasbro
Tom Kalil,, The White House
Tim Matt,, Siebe
Jean Scholtz,, DARPA
Ted Selker,, IBM
Randy Sweeney,, Kraft Foods
Jim Waldo,, Sun Microsystems
Kevin Kelly,, WIRED magazine

Pretty fun list of people so far. This is a good sign. Given what I've learned
of Prof. Gershenfeld's Hyphos wireless embedded net at the Media Lab, I could
even call myself excited. Here's the official blurb:

> The PC monolith will break down, and concentrated "core" elements of
> computing and communication, sensors and actuators, will become
> embeddable in almost everything. The "jellybean" processors that
> currently pervade nearly every appliance, yet are utterly isolated,
> will be connectable through a wealth of emergent capillaries
> sprouting from the internet. Technologies will be produced that are
> ultra-cheap, ultra low-power, and radically different from today's
> chip-and-pc-board variety (think: printable circuits, wind-up
> electronics, wearable networks powered by walking or breathing, even
> edible circuitry). Ingredients like these will form the foundation of
> a vastly extended network of things that are very different from
> PC's. Within ten years, a billion people on line will be joined by a
> trillion things with embedded networks.

> The goal of this workshop is to convene a limited number of
> leading engineers and researchers from a wide cross section of
> academia, industry, and government to discuss critical challenges in
> developing and deploying embedded intelligence over a wide range of
> applications. These are "out of the box" systems in every way, shape,
> and form. They demand big, bold, maverick thinking.

> This 2.5-day meeting will consist of invited talks, refereed
> papers, and work-in-progress reports. Informal mingling opportunities
> include an Open House at the Media Lab in conjunction with the Things
> That Think consortium meeting, and plenty of schmooze time.

No registration fees. Based on my last conversation with Kalil, there's reason
to believe there's money following this area. Jean, I believe, is the program
manager for Localizers. I'm beginning to think it's time for the Munchkins
Paper (TM) to finally be written: IP Considered Harmful :-)

> Topics of interest include, but are not limited to:

> Applications in unusual domains: toys, appliances,
> cars, human implants, domestic, rural, outdoor, undersea...
> Capillary network architectures (Bluetooth, IrDA, PLC, etc) Software
> systems to make these systems work Case studies, and cost-benefit
> analyses New interface paradigms Self-healing and self-assembling
> systems Drastic scaling issues Secure communications Emerging
> standards

> Awards will be given for the best paper and best student
> paper at the conference.

> The USENIX student stipend program covers travel and hotel
> to enable full-time students to attend.

On the other hand, research agendas may not top their list of exciting
prospects. A derivation of the PagerNet proposal might, though -- and
defintely SensOS.

> Attendance will be based on submission of an extended abstract as
> evaluated by the program committee. This should describe original
> work concerning the design, implementation, and real application of
> embedded systems. We are not looking for tweaks to Linux, or stuffing
> WinCE palmtops into toys. Rather, we are seeking radical new
> architectures, exceptionally promising prototypes, enlightening case
> studies. The abstract should convince the reviewers that a good paper
> and 20-minute talk will result. Identify what has been accomplished,
> why it is significant, and compare it with relevant work in the
> field. Include references, illustrations, and performance data. Be
> incisive and cogent.

I'm stoked. I'm disappointed that the Workshop was announced before Christmas,
but my Social Intelligence networks failed to hoover this item up until today.
Still, two days is about as much time as I would have imagined spending anyway



Rohit Khare -- UC Irvine -- 4K Associates -- +1-(626) 806-7574 --