For Wednesday morning, I decided to attend the Distributed Teams talks as it seemed to be the closest match to my interests around helping organizations work together, and I wasn't disappointed.
Capturing and Supporting Contexts for Scientific Data Sharing via the Biological Sciences Collaboratory by George Chin, Carina Lansing of the Pacific Northwest National Laboratory.
PNNL have created a portal through which their ~100 biological sciences people can interact. They talked about a number of the organizational principles behind building the tool and what they learned as a result. One of my favorite concepts came: context. This time it was framed as a way to look at or think about data sets: general properties, experimental properties, physical organization, project organization, scientific organization, user community, experimental process, tasks, data provenance, integration from and with other sources, and analysis & interpretation. These context can be used to peer into the data sets from many different perspectives. I don't know that they have gotten all the way to this point, but the general idea seems to be there. Interestingly, this sounds very much like the KM Chicago talk I attended last night that had to do with building a metadata model for an entire organization in order to help them drive to new levels of effectiveness.
The tool also has electronic laboratory notebook capabilities and the interesting ability to look at large data sets and how they change over time. Essentially, one could set up a search against a data set to watch where and how it changes. And they intend to continue development and expansion of its capabilities into dissemination / publishing; peer review; confidence, reliability and integrity; protecting research (IP); and classification (for national security need).
Meeting Central: Making Distributed Meetings More Effective by Nicole Yankelovich, William Walker, Patricia Roberts, Mike Wessler, Jonathan Kaplan, Joe Provino of Sun Microsystems (paper pdf).
This team from Sun have developed a new application that enhances teleconferences with graphical indicators of who is online, speaking, muted and even who is having trouble. There were motivated by research they did that showed none of the current technologies met the needs of their internal users associated with distributed meetings. The tool works directly with their VOIP lines to retrieve identification information. It actually calls the participants, rather than having participants dial into the system. Along with the graphic indicators mentioned above, it has the standard shared-desktop that is familiar to many of these applications. It also has a cool capability to transfer the call from your home line to a mobile line or other location, when needed. Can't wait to see it available as a product!
Amazingly, David Weinberger has just written about something quite similar with respect to VOIP. If phones didn't already exist...
Why assume that phone calls have to be audio only? That's an artifact of the current infrastructure. If you were starting from scratch and didn't have lines that only knew how to carry sound but could carry any type of bit you'd build something far different. It'd integrate with other applications on your phone device. It'd know who's calling from where and spin up a web page to show you the relevant information. It'd link to everything the Net knows. It might assume an open mike -- not a two-person conversation -- as the default. It wouldn't assume that each phone call is essentially separate from every other; it would assume they can cluster into threads the way emails do. Or perhaps it'd find a more natural lumping than threads, including joining with other types of communication. It'd provide IM as a backchannel for multiperson phone calls. It'd give you visual bits as well as audio ones. It would do things that no one can predict because the genius of the market hasn't yet invented them. But it is in the process.
In-Group/Out-Group Effects in Distributed Teams: An Experimental Simulation by Nathan Bos, Judith s. Olson, Arik Cheshin, Ning Nan of the Collaboratory for Research on Electronic Work at U of Michigan; and N. Sadat Shami at Cornell.
This work looked at the problem of the "unpopular kids," which in research terms is the Out-Group, particularly when those unpopular kids are coworkers who are not co-located with the rest of the team. They devised a game to test some ideas around the impact of this idea. The game required participants to ask for help from one another, and one group was located in a room while others were only online. In their results, they observed that the two groups essentially had to form their own in-groups. The collocated group worked together by talking and the telecommuters worked together by sending messages. (They sent messages to everyone at the start, but then only interacted with the people who wrote back, creating a quick disparity between the two groups.)
They found the telecommuters and collocators got to success via different pathways. They also noted that those who failed amongst telecommuters failed miserably because they had no one to rely upon for help. The telecommute group also reported a better time of concentrating. The limitation I see with this particular experiment was that the telecommute group had essentially everything it needed to succeed at the game, whereas most cases of distributed teams find the out-group needing critical information from people in the in-group.
[Thanks to the back channel for helping me understand some of the unfamiliar terms here.]