M2C, The Materials MicroCharacterization Collaboratory

Dire chimera's conquest was enjoined;
The mingled monster, of no mortal kind;
Behind a dragon's firey tail was spread;
A goat's rough body bore a lion's head;
Her pinchy nostrils flaky flames expire;
Her gaping throat emits infernal fire.

--Homer, Iliad

Figure 19. M2C Collaboratory Logo

At least seven Industrial Partners participate technologically and financially in the M2C Collaboratory. Industrial Partners include Gatan Inc., R.J.Lee Group, EMiSPEC Systems Inc., Philips Electronic Instruments, Hitachi Scientific Instruments, Inc., JEOL USA Inc., and Graham Technology Solutions. None of these partners have a strong graphical or informational presence on the M2C Collaboratory's five TPM sites.

Together, the M2C Collaboratory's five TPM sites have $50 million in rare and expensive microscopy instruments, with key pieces hardwired to the network, and hundreds of highly trained scientists and technicians with tens of thousands of hours of specialized scientific experience at each others' disposal. None of the labs do exactly the same thing, and each has specialized and unique instruments that, when brought together as M2C, do the larger job at each individual lab.

The "hand" of the M2C Collaboratory seems to be at the Argonne National Laboratory, http://tpm.amc.anl.gov  (Zaluzec 1999), from where the coordinated web video telepresence and an aggressive public relations program originates, and the M2C's most public personality, Nestor Zaluzec, is located. The unofficial "head" is at Oak Ridge National Laboratory (Wright 1999) where the information science aspect of the project is centered.

Zaluzec arranged for two video cassettes (Argonne 1998, Dept. of Commerce 1998a), and a CD-ROM (Dept. of Commerce 1998b) about M2C and DOE 2000 to be sent to the researcher as soon as consent to enter the Argonne TPM site was granted. The researcher was also advised of a recently published "hot" article about the M2C Collaboratory (Kling 1998), which provides a nice introduction to the project, but which was not discovered during Phase One's library research. The multimedia materials are intended for educational use, are distributed freely to the public, and are available by request to the U.S. Department of Commerce NIST Public & Business Affairs Office in Gaithersberg, MD.

The professionally produced CD-ROM and video cassettes present the story of an exciting collaboration between scientists at NIST, Argonne, and Texas Instruments (TI) during which a microscopic problem that developed during silicon chip production at the TI manufacturing plant in Texas is resolved online in a M2C Collaboratory session. The TI scientist prepared and sent the problem chip to Argonne. Argonne positioned the chip in one of its microscopes, and opened a collaboratory session with TI. The TI scientist could see through the eyes of and manipulate the microscope, and collaborated with the Argonne microscopist to identify the problem and devise the solution.

But, full color, multimedia glitz and promising public relations aside, the actual experience of M2C Collaboratory's virtual facility is a hard-on-the-brain cacophony of overcrowded information presented on a battleship gray webpage awash in a sea of acronyms. Each Argonne TPM web page is chunked into five or more scrollable frames, all crowded onto one computer screen. One quick click on the wrong acronym switches the participant, without fanfare, from one lab to the next, one instrument to another, from one side of the country to the other. This arrangement makes it difficult to achieve and maintain orientation. Nevertheless, the promise that anyone, anywhere in the world, can log on to the public Internet, go to the Argonne TPM web site, and fine focus a million dollar microscope located at any one of five National Laboratories around the country, overshadows complaints about the cacophony of the interface. This potential is a powerful feeling, and provides a taste of what the future may hold for virtual scientists.

Consent to enter the ORNL, U of I, and NIST TPM sites was also received. Consent to enter the LBNL TPM site was denied because the site would be inactive during the term of this research, and the primary contact at LBNL was out of the country. The testsite experience, then, became an exploration of four Collaboratories, as one.

TPM : TelePresence Microscopy

The Argonne TPM site is the control center of the M2C's telepresence. It coordinates constant video feed between and among the five TPM sites using "push" technology. It does not support audio transmission, so the experience is all eyes, no ears. The video transmission rate over a simple dialup connection to the web is three frames per second and begins as soon as the website is accessed. Three frames per second delivers jerky video: more a slow steady series of independent snapshots than a flowing stream. The connection, however, is constant and consistent if the client system is not running superfluous programs, particularly anything from the Microsoft family of software, the AutoSave functions of which consistently cause crashes and session interruptions. Argonne recommends a high speed dedicated connection for full effect, and use of the Netscape Navigator web browser.

Whether through a dedicated connection or by Internet dialup, maintaining a connection to M2C's video feed requires lots of RAM, or Random Access Memory, the client "desktop" space available during any computer session. For ease of this research, a personal computer dedicated exclusively to the collaboratory session was necessary (additional systems configuration and requirements are discussed later in this chapter.)

Argonne's TPM website provides multiple, simultaneous live or recorded video transmissions from itself and any ONE of the other four TPM sites, with as many as six separate video transmissions available on a single webpage. Switching between the video views and between TPM sites is as quick as a click. The web telepresence uses a point-to-point video feed rather than a multi-point videocast, as it would if all five sites were sending and receiving video simultaneously (which TPM facilitates on special occasion using CuSeeMe teleconferencing software). Participants are able to switch between and control the views from the microscopes and the macroscopes (room view video cameras) at any of the sites. Split screen "collaboratory sessions" allow the participant to view microscope and macroscope video transmissions from two cameras at each of two sites simultaneously for a sort of "four-eyed" experience. Everyone accessing the TPM website sees the same video feed, and anyone can change which video is seen. The host-site may override and disable a remote controller and any change in video view changes the view for everyone. A warning is posted that if the view suddenly "changes" it is because someone else who is accessing the site has switched views. No information is provided about who else, or how many others, are accessing the TPM sites, so it feels as if only you, and the people you see on the macroscope view from the lab (if there are any), are present, when in fact thousands of others might be.

To further confound the confusion about who, or how many, are seeing what view from where, the number of possible simultaneous video feeds on a single screen is six, making a sort of six-eyed experience with each eye seeing something different. Getting oriented to what is on the screen is difficult and confusing, and often, as soon as orientation is gained, the "view" suddenly changes.

The Argonne lab is the video "headquarters" of the M2C Collaboratory, and must be included as a partner in any two-lab collaboratory session. The Argonne setup does not allow direct connection between two remote labs, only between itself and one other lab, or any other lab and itself. The same is true of the other sites, which allow a view from themselves and Argonne, or from Argonne and any one other site. Accessing any one of the labs by itself is also possible, either from a connection at Argonne, or by logging on to that TPM's separate URL.

For instance, a collaboratory session between the Oak Ridge Lab and the Berkeley Lab is not supported; however two sessions can be set up simultaneously, as between Argonne and Oak Ridge, and Argonne and Berkeley, and participants can switch between these sessions, requiring orientation to as many as twelve video feeds. Collaboratory sessions are preprogrammed and executable with a one click switch to "collaboratory mode." Remote users may "join" an ongoing collaboratory session, and will receive microscopic and macroscopic video from two labs on a single screen (see "collaboratory mode" later in this chapter). In other words, participants may be visually "in" one of the labs alone, or, in collaboratory mode, be "in" two labs at once. Participants can even be in multiple collaboratory sessions, conceivably four at once.

For a scientist at one of the labs, or for one of the M2C Industrial Partners, or someone from the wider Internet, to conduct an experiment using one of the instruments at any of the five TPM sites, the materials sample has to be prepared and sent to the lab, and a technician or scientist has to receive the sample and load it into the instrument. This setup, then, allows the remote scientists to have their eyes at the remote site, one on the lens of the microscope, and the other viewing the room, with their hands on the remote control buttons for the microscope in a separate frame on the M2C webpage. They can see through and manipulate the microscope, but they cannot move the sample at all, or talk to the remote scientist without a subsidiary telephone connection. A "snapshot" of the microscope's results can be viewed either as real-time online video, or the results can be captured and posted to a collaboratory notebook (the notebook configuration is discussed in detail later in this chapter.)

It is possible and practicable to connect to and enjoy the full function of the M2C Collaboratory and participate in TPM activity with a 14.4kps dialup ISP connection to the Internet from a 486-66 mhz PC with 16 mb RAM and a standard video monitor. It is slow, but possible. A frames-capable version of Netscape NavigatorÒ (2.0 or higher) is required to receive video; otherwise, no special software or unusual configuration is required. Microsoft's Internet ExplorerÒ browser does not support the file type used for video transmission. This research used a 28.8 kps dialup Internet connection from a 200 mhz Pentium II with 128 mg RAM and 2 mb Video RAM, as well as a 166 mhz Pentium laptop with 16 mb RAM and 1 mb Video Ram connected by dialup via a 56.6 kps modem.

Only one area of each TPM site has restricted access: the "Online Control" of the hardwired instrumentation. Online control of the instruments must be justified and arranged ahead of time. As U.S. National Laboratories, however, each of the TPM sites is required to make their instrumentation available to qualified outside researchers. A form for requesting instrument use is online at each site. However, users may view others' use of the instruments, or read the contents of the public notebooks, (password protected private notebooks are also set up) and may remotely manipulate the site's macroscopes without password access.

Video from and manipulation of the room-view cameras is provided around the clock, whether anyone is in the lab, or not. However, video feed is often not "live" in that it may be a prerecorded transmission intended for demonstration purposes. It is often difficult to distinguish whether the video is live or prerecorded. Of course, macroscope and microscope views that are prerecorded have the remote control options disabled.

Each TPM site hosts shared public and private notebooks. The notebook is a shared, online document space in which participants may post pages containing text or still graphic shots captured from the microscopes. Only one person can have realtime control of any given notebook page; so, chat-like communication is not possible. Private notebooks are password protected and reserved for specific experiments, which may be of a proprietary nature, as with a collaboration with Texas Instruments or one of the Industrial Partners. Public notebooks allow real-time addition of pages via web forms technology. Each notebook page has an email-like header identifying the author, subject, etc. While there is a provision for searching the subject line of notebook pages, there is no provision for full text search, nor for annotating or linking among notebook pages, or among notebooks.

The Argonne TPM site hosts a data archive where programs, software patches, and activity logs are stored and shared via file transfer protocol. The site hosts an "Ask-the-Microscopist" service, and provides information about setting up middle/high school experiments. There is a questionnaire and comments form, and access to join or read the archives of a six-year-old distributed email list. The researcher was granted access for research in the archives of the M2C Collaboratory email list, an opportunity which will provide data for expansive studies of other aspects of the collaboratory ecology, but which was not undertaken in this study.

The site design relies heavily on web frames (dividing the screen into multiple, smaller windows) to organize and present these many options. The interface is cacophonous and confusing. The framed sections quickly become quite small and are packed with options, most of which involve arcane acronyms that may be confusing to the novice user. The confusion is compounded because each of the five laboratories has its own specific set of acronyms along with an individual name and domain, and one click can mean a switch between labs and instruments without noticeable change in layout or design of the web interface. At certain times, the M2C Collaboratory's TPM sites feels as if it is five labs in one, and at other times, five individual labs.

This study monitored the M2C Collaboratory through the Argonne TPM site at odd times and for varying durations and intensities, on at least a weekly basis, for ten months from October 1998 until July 1999, for an estimated twenty hours total observation time. It wasn't until three months into the experience that the "who's on first" feeling of disorientation was relieved, and comfortable orientation to the larger environment was achieved. The final confounding variable is that contact information from each of the sites leads back to Argonne, making the Argonne TPM site appear as if it is the center of the Materials MicroCharacterization Collaboratory. But M2C may also feel "centered" at any of the other TPM sites when primary access is made through that lab's distinct URL. M2C is actually a megacollaboratory encompassing five individual but intertwined collaboratory sites: ANL, LBNL, NIST, ORNL, and U of I.

During the ten months of this study's immersion, the main microscope at TPM Argonne was frequently offline, and two of its seven microscope lenses were under repair. No public collaborative experiments were conducted. Many times a prerecorded video of collaboratory activities was webcast and it looked "as if" there was live lab activity, when in fact it was not real-time activity at all, but prerecorded video.

Because the macroscope controls were accessible even when no one seemed to be in the lab, users are free to "poke around," panning the cameras, and focusing in and out on pieces of equipment and researcher desk space to get a sense of the place. Because the transmission rate is three frames per second, however, and because the views could be changed without notice by some unknown, other visitor, sometimes someone would just "appear" in a chair in the laboratory, and just as quickly "disappear" between frames. It was not immediately possible to tell whether this appearance was real, or recorded, and often from which lab it originated. In fact, it was generally difficult to tell what was "real" and what was recorded, what was live, and what was contrived, during most of the M2C sessions. At no time during the ten month immersion did any collaboratory participants make contact with the researcher while in the collaboratory. No response was posted to the researcher's entry in the public notebook. The M2C Collaboratory's video telepresence provided a sense of, but no real connection to, other humans.

Because of the specialized, scientific focus of the M2C, an understanding of instrument type and function would help with orientation, and for educational purposes an elementary overview of the equipment and the work it is used for would be helpful. Much of the researcher time was spent learning which piece of equipment did what sort of job. Without this specialized knowledge it is virtually impossible to get a comfortable sense of the work that takes place in the M2C.

The Argonne TPM public notebook contains 125 entries, dating from March 1997. Most of the posts are brief messages concerned with startup technicalities and configuration tests for various graphic settings. Again, this was no place for beginners as none of the messages made an attempt to explain either technicalities or acronyms. There are a few series of messages (indicated by the subject line of the notebook entry), which are mostly single frame graphic files that appear to have been produced on a microscope during remote collaboration with a microscopist. None of the notebook pages that containing graphics provided more than a brief, acronym-laded description of the shot; there was no discussion of the contents of the graphic file, nor any sort of formal diagnosis or discussion of the materials problem.

The M2C's digital library (although the word "library" is not used) is a separate collection of internal and external web links maintained by the microscopist/webmaster at Argonne. The link to the digital library (called "General Info about TelePresence Microscopy") is not prominently displayed but, once located, transports the user to a separate web address at http://www.amc.anl.gov , home of the Microscopy and Microanalysis WWW Server. The server offers a collection of project papers, proceedings, and reports aimed at general education about the M2C concept and the science of materials microscopy. None of the documents were for specific project support nor were they about ongoing experiments. They seemed intended for general visitor information rather than to support collaboratory activity. The "General Information" site also points to data archives that include the site's distributed email list archive, and links to related microscopy societies and other external WWW resources. There were no preprint papers, and no peer review postings. It felt as if the M2C Collaboratory's main purpose was the exchange of video and captured shots from the instruments. Very little attention is paid to the intercommunication or inculturation of participants.

With all the CIRAL criteria in place and functioning (Table 7), the Argonne TPM site of the M2C Collaboratory proves the collaboratory exists as it was philosophically, intellectually, and instrumentally described by Wulf (1988), Lederberg and Uncapher (1989), and NRC (1993). It is not possible to say M2C was the first or is the only full implementation of the collaboratory, but it is possible to say that the Argonne TPM site of the M2C Collaboratory appears to have been fully functional in March 1997, when its public notebook went online.

NIST TelePresence Microscopy Site

At first, there is no apparent difference in the look and feel of the Argonne TPM and the National Institute of Standards and Technology (NIST) TPM site, and for a while the researcher was not aware they are two distinct TPM sites. Even though the online address, or URL, is different, the interface and design are nearly identical, as if the Argonne TPM webpages were mirrored on the NIST site at http://scanner.cme.nist.gov . Only the acronyms seemed to change. But, on closer inspection, the Argonne TPM site showed 41,566 visitors logged on since 1997, while the NIST site logged 1801 visitors. The microscope video feed from each site is different (although this might only be apparent to a microscopist). All other options seemed identical.

Achieving collaboratory mode between the Argonne and NIST TPL sites is a one-click operation that brings both sites together onto a single screen. Micro- and macroscopic video views from each site are framed above other frames that contain video control, comments, and instrument controls, for a total of seven frames on a single webpage. Similar sessions can be configured between Argonne and ORNL, Argonne and U of I, or Argonne and LBNL.

Figure 20 is a screenshot of a TelePresence Microscopy session between Argonne National Laboratory (ANL) and National Institute of Standards and Technology (NIST) on May 16, 1999, as seen from an Internet connection to the website at http://tpm.amc.anl.gov/TPMLVCollabNIST.html .

The upper left frame is a lens shot from the Advanced Analytical Electron Microscope (AAEM) at Argonne in Illinois. The upper right frame is a shot of the NammMetrology SEM Microscope at NIST in Maryland. The top center shot is a macroscope (room) view of the Argonne lab, and the middle center shot is the macroscope (room) view of the NIST lab. The lower left frame has buttons to change between the instrument or cameras at either site. The lower center frame contains links to switch between collaboratory sessions and to other web- based resources. The lower right frame is the online instrument controls, which, in this shot, are offline.

Figure 20. Screenshot of M2C TelePresence Microscopy Session

Consent from ORNL's TPM gatekeeper suggested the researcher access the M2C from http://tpm.amc.anl.gov/MMC/ , yet another portal to the site. This access point provides a thorough overview of the M2C Collaboratory and includes an introduction and background information, goals, tools, status, opportunities for involvement, expansion plans, partners, visions, publications, workshops, and images. It includes a form to request use of resources or information at/from any of the TPM sites, email access to TPM gatekeepers, an overview of research programs, and connection to the project's steering committee. While all this text-based information helps make sense of the science of microscopy, and the notion of the collaboratory generally, it does not aid visual orientation to the TPM experience.

The M2C experience has a sort of "virtual" feeling of being there, somewhere, within and among these five labs, but in eyes only, and with a lingering and somewhat creepy sense that others MAY be around, but are hidden. Because there was no human interaction, either via audio transmission or real-time chat, and because it was difficult to determine whether the video feed was live, or prerecorded, and also difficult to orient to which lab one was connected, the M2C experience was generally disorienting and produced a feeling of disconnected connection, of being alone online with unknown virtual others.

Conclusion