I attended an interesting seminar from Professor Rodney Brooks, director of the Computer Science and AI Lab at MIT yesterday afternoon. He’s visiting Australia as a keynote speaker for the ICT Outlook Forum (which I attended last year in Canberra), and stopped by to spend the day at the CSIRO ICT Centre. The main gist of his talk was about exponentials, and how exponential patterns are the most influential indicators of future technology trends. Of course, the most well known of such trends is Moore’s Law, and predictably, Brooks hammered on this quite a bit. He did, however, also come up with some interesting statistics: There are approximately 10^16 ants in the world, and 3×10^16 grains of rice grown per year; in comparison, in 2003 we produced 10^18 silicon transistors – i.e. 100 transistors for ever ant in the world, and 33 for every grain of rice. This was his way of illustrating the significance of exponential growth, and I think it’s fair to say he made his point well.

Other potential exponentials Brooks identified (i.e. trends that look suspiciously like the beginning of exponential curves) included multi-core processors, such as the CELL processors. In the research labs, we’re already seeing 64 cores on a single chip – is this the beginning of a new exponential? If so, IMHO it will almost certainly change the way we develop software quite dramatically. I think it’s more than fair to say that at this point, most software is not designed with high-levels of parallelism in mind. Given this, and the different caching model in processors like CELL, there is certainly a requirement for further research and development into compilers that can both hide the complexity of multi-core processors from software developers and deal with a substantially different caching model to current processors. More interestingly, what kinds of things will be possible with the additional computing power of such chips? What can you do when you’re wrist-watch or your microwave is as powerful as a current-day supercomputer? Of course, this is nothing more than Moore’s law continued, but with greater parallelism perhaps substituting for raw computing power in a single CPU core.

The other big emerging exponential that Brooks seems fixated on is that of personal storage. The iPod is seemingly the poster child of this exponential for Brooks: he cites the fact that (US) $400 bought you a 10Gb iPod in 2003, a 20Gb iPod in 2004, and the equivalent of about a 40Gb iPod in 2005. If this trend continues, we’ll see 40 petabyte iPods within 20 years. Of course, this then raises the very obvious question of what the heck would we do with such storage capacity? It’s pretty much the analogue of what the heck can we do with 10 gigabit network connections that’s appealing to a wide range of people (sure, we can always think of some special cases that already require massive amounts of bandwidth or storage, but what about mass-market applications?).

Brooks posed the magnitude of such storage space in terms of books, photos and videos – By 2009, we’ll theoretically be able to walk around with 1 million books loaded (as text) on an iPod in our pocket. What the heck does this really mean? None of us can possibly read 1 million books in our lifetime: even if we live to 100, that makes about 30 books *per day*, every day of our life. So, what does it mean to be able to carry around the text of 1 million books in our pocket? Or the whole Library of Congress by about 2013 (for some reason, Americans are obsessed with this as an example). For starters, it seems to me to be almost at odds with the ‘always connected’ vision of the future. Who needs to be always connected if you can cache so much data? Sure, there will always be some data you want real-time access to, but it would seem that a lot of the existing web-style data could easily be stored and sync’ed at regular intervals on some massive personal storage device. Blogs, news, code, reference information even DNS data: all this stuff is cacheable. I guess the whole podcasting phenomenon is an example of this mode of interacting with data.

And what about beyond books? Well, Brooks’ calculations have it that we’ll be able to store something akin to every movie ever made (or close to it) on our portable personal storage device by about 2025. Again, it’s a trite example, since we could never hope to (or want to) watch so many bad movies. But again, it hammers home the question of what we can imagine using such space for. Will we walk around with the ID of every RFID tag ever assigned? What about a complete copy of the global DNS and a big chunk of the Googlzon search index? Funnily enough, it’ll be a very long time before we’re able to keep track of each silicon transistor (all 10^18 of them per year) on personal storage, but then I can’t say I can imagine anyone being overly disappointed about that! Maybe we’ll be sending multimedia or holographic messages to each other, and we’ll be able to keep a copy of everything we’ve ever communicated. What about a recording or transcription of everything we’ve ever done or said, image and video data of every place we’ve ever been, the face and voice of every person we’ve ever met? Well, that’s just starting to sound like Gordon Bell’s My Life Bits project. If indeed we do head down this path (which seems plausible though not inevitable) a more important question that arises is: how the heck to we deal with such an overload of data and information with our limited cognitive abilities? What kind of tools do we need to develop to help people wade through the swamp of every piece of information they’ve ever interacted with? Certainly, we’ll need to be able to tailor the information that is retrieved and the way it is presented according to the current context if people are to have any hope of using any of that mountain of data in their pocket. Unsurprisingly, that’s exactly where my team’s research is focussed.

Assuming we can make some progress on the retrieval and delivery of information from your personal data swamp, how can you imagine using personal, portable storage that is for most intents and purposes limitless?

Shell and I completed another Hawkesbury River Classic training paddle on Saturday, and in doing so we set a new personal record for the furthest distance we’ve paddled so far in a single, unbroken session. As you can guess from the post title, the paddle was from Windsor to Sackville, which is actually the first leg of the Hawkesbury Classic. Windsor is a good 30-40 minutes drive from home, so it was an early start to get up, ready, packed and out to Windsor by 7:30am.

After dropping Shell, our kayak and all our gear off in Windsor, I had a quick car shuffle (which included a ride on the Sackville car ferry) to have the car waiting for us at Sackville when we finished. By the time we’d all sorted out the car shuffling, got kitted up in PFDs, registered, attached race numbers, had a quick briefing etc. it was about 9:30am when we got on the water. Again, we were certainly amongst the more recreational paddlers on the day – in fact there was really only one other boat there that wasn’t a racing kayak (that kayak had a home made keel attached to the bottom to keep it tracking well).

Without going into the details of nearly four hours of continuous paddling (I think we finished in about 3 hours 50 minutes in the end with two approx. 5 minute breaks), it was a really enjoyable paddle, although the final few kms were tough. We set off from the starting line at breakneck speed, with a stroke rate that was way above what we are able to sustain for anything more than a few minutes. (Shell later confided that she’d kinda got caught up in the moment and was seeing if she could match the stroke rate of the more hardcore marathon paddlers around us). After our flying start, we eased into a more regular rhythm. The first 25 or so kms went by pretty happily – we made good time, and had the opportunity to chat with a few other paddlers along the way and take in the varying scenery we were paddling through. The final few kilometres were quite a bit harder, at least for me, but eventually we caught sight of the Sackville ferry and knew we were going to make it. Of course, shortly before we finally reached Sackville, we were passed by some of the more enthusiastic paddlers heading back to Windsor – apparently 32km on a Saturday morning isn’t enough for some people!

One thing that made the paddle a lot easier was our new CamelBak drinking bladders. With these fitted behind our seats and the drinking tube tucked into our PFD, it was a lot easier to have frequent drinks without having to stop paddling to unstrap and unscrew a bottle each time. We still have to work out the correct positioning of the tube (sometimes we still needed to miss a paddle stroke to position the tube correctly for drinking), but it’s a vast improvement over using our Sigg bottles. Surprisingly, they were quite cheap too (well, cheap compared with most of the other CamelBak systems), because they don’t come with a harness etc. We reckoned we’d be better off without having to carry the weight on our backs, and after Saturday’s paddle, I think that was an excellent decision!

Of course, we’re back at it again this Saturday – this time from Sackville to Wiseman’s Ferry, which is I think a shade over 40km. I think we’ll need to buy some kayaking gloves before then – even with calluses on my hands, they are still starting to blister by the end of each paddle. Not to mention the dry and rubbed skin in the webbing between the thumb and forefinger.

It seems NICTA and the WA government think the CSIRO ICT Centre/Queensland Government E-Health Research Centre is a great idea – well, good enough to emulate for their own recently announced partnership, the International Centre for Health Care Solutions, also known as e-Med. Alarm bells start ringing, however, when the article states that e-Med will focus on “futuristic virtual surgical training” and “portable medical devices”, areas in which CSIRO has previously done significant work. Hopefully politics don’t get in the way of collaboration between eHRC and E-Med (and indeed other groups) – there are a huge number of critical health problems to be solved and it would be a great shame to see duplication of effort for the sake of political point scoring.

PC World has published an article (PC World | CSIRO launches flying datacentre) on our recently completed 3-year research project with Boeing (USA) around developing new technology for the RAAF Wedgetail airborne early warning and control (AWACS/AEW&C) aircraft.

With typical journalistic flair, the story has been blown up a little: I’m not sure I’d quite agree that the technology we developed “has also been commercialized for sale to appropriate customers”, nor have there been 20 scientists and engineers working on it (well, maybe close to that number contributed, but there certainly weren’t anywhere near that number of people working full-time for 3 years, as might be inferred from the article) but the important parts are there.

The focus of my team’s contribution has been in intelligent information delivery: how do we prevent air surveillance operators from being overloaded with information, while still ensuring that they aren’t deprived of and don’t overlook any important information? Initial investigations by Robert Tot of current air surveillance operators at the RAAF Williamtown airbase allowed us to observe operators in action to understand the information they use to perform their work tasks. Interviews and observations also highlighted several issues: 1) Operators have to manually integrate information from a number of different sources to perform their job. This can include having to physically move to a different computer terminal (e.g. to access civilian flight plan information). 2) Displaying all of the available information all of the time is infeasible because the display becomes too cluttered.

Our approach to alleviating these and related problems was to create an adaptive graphical user interface that tailors the information displayed at any point in time and how that information is presented according to the operator’s current task and role. Based on this context, the relevant information required by the operator is planned, gathered and delivered using Myriad, our java-based platform for contextualised information delivery.

At the core of Myriad is our Virtual Document Planner (VDP), a goal-decomposition planning engine that, when configured using a set of plans, produces structured representations of content to be delivered that is specific to the current interaction context (which includes who the information is being delivered to, what task they are currently trying to perform, what environment they will view the information in, what information they have previously been presented with etc.).

The structured representation of content produced by the VDP explicitly models the role of each fragment of information to be delivered, through making explicit the rhetorical relations between each piece of information. We can then reason about the content, based on its structure, in deciding which information should be presented and how, based on whatever constraints might apply (e.g. temporal or screen-space constraints).

In order to infer the current (and future) tasks being performed by an operator, our Operator GUI provided a constant stream of user actions to a Task Parsing module, which based on a grammatical model of the operator’s possible tasks, makes statistical predictions of: what task is currently being performed, what task is likely to be performed next, and what information is required by the operator to perform these tasks. This information allows Myriad to plan the delivery of information proactively, meaning that operators shouldn’t need to request information; instead they should find that information is discretely made available to them as they require it.

Of course, the proactive delivery of information risks overloading or distracting the operator, who may be deeply engaged in other current tasks. To avoid unwanted distraction or disorientation, we were very careful to provide new information by discretely displaying a notification of information availability, rather than immediately providing the information itself on screen. In this way, we leave the human operator in control to choose if and when the information is required in order to complete their tasks.

The project has required me to develop our Myriad platform to support the delivery of textual, graphical and spatial information. In addition, I have been responsible for the development of the intelligent, adaptive Graphical User Interface. The GUI is based around the excellent OpenMap framework from BBN for the display of spatial information. In addition, a desktop-like workspace has been created where more verbose information could be displayed (either linked to objects visible on the map, or provided as non-spatial data). To allow the GUI to be completely controlled from Myriad, I created a flexible and extensible command-line API (using the BeanShell Java source interpreter), through which information can be added, displayed, hidden, modified, highlighted etc. on both the map and workspace displays with commands sent to specific GUI channel listeners over message-passing middleware.

I’ve just heard from HCSNet that ten travel bursaries of $500 towards travel and accommodation costs are available to PhD students from outside metropolitan Sydney who wish to attend the NICTA-HCSNet Multimodal User Interaction Workshop, to be held at the Australian Technology Park, Redfern, Sydney, on September 13-14th, 2005.

The workshop includes two invited talks from internationally-recognised researchers in multimodality: Professor Sharon Oviatt from the Oregon Health and Science University, and Professor Francis Quek from the Virginia Polytechnic Institute and State University.

In the interests of information sharing, it is a condition of receipt of a travel bursary that the student should provide a poster describing their current research project.

The closing date for applications for bursaries is Friday August 26th 2005. Those interested should send an email to the HCSNet Convenor (Professor Robert Dale) at rdale at ics.mq.edu.au.

If you’re interested in knowing a little more about the CSIRO ICT Centre (where I work) take a look at this profile of our director, Alex Zelinsky in the Sydney Morning Herald. I certainly don’t mind being counted among “200 of the nation’s brightest computing and engineering researchers”

I also like Alex’s philosophy that “Talent attracts talent”. Google is of course the most extreme example of this right now, to the point that I’ve read several recent articles lamenting that all other tech companies in the US are currently suffering a brain drain to the search behemoth.

While we’re not quite at the same scale as Google, if you feel like joining our local pool of talent, you’ve only got one more day to apply for the software engineering position that closes tomorrow.

If you’re interested in speech, language or sonics you should consider joining HCSNet, the Australian Research Council Research Network in Human Communication Science.

HCSNet aims to bring together researchers and students through workshops, conferences, and a variety of collaboration schemes in order to explore the boundaries of disciplines that encompass human communication. As a guide, this includes fields as diverse as psychology, computing, linguistics, engineering, philosophy and music.

Being a participant gives you access to an increasing number of funding programs (including funding for running interdisciplinary workshops and seminars) and events that are run under the auspices of the network. There’s some really good stuff coming up, including the NICTA/HCSNet MultiModel User Interaction Workshop (free registration thanks to HCSNet funding!), so join up and you’ll get the weekly HCSNet newsletter that will keep you informed …

I’ve spent the past two days at the inaugural CSIRO Scientific Development Forum (CSIRO access only, sorry people!), a gathering of early-career scientists and post-doctoral fellows from across CSIRO’s many diverse research areas. The focus was on providing insight and tools to help us develop as the future leaders of science in CSIRO and the local and international community.

Much of the two days was spent in discussions about our various diverse research areas. This was fascinating, given the breadth of our interests – spanning everything from molecular science, genetics, food technology, animal cloning, virology, astrophysics, precision optics and superconductor physics, petroleum research, ICT research, statistics, nutrition and a bunch of other areas. Despite such breadth, there were many interesting insights shared about how to take control of our own careers and personal development.

In addition, we heard from several high-profile speakers, including Federation Fellow Professor Calum Drummond, who talked about his career. One thing I certainly took away from his talk, apart from his “can-do” attitude, was his emphasis on the importance of balancing focus on work with family life. Dr David Mitchell spoke about entrepreneurship and his experience in several start-up companies. We also workshopped and brainstormed some interesting ideas around identifying business ideas that might satisfy existing market gaps and market needs across a range of industries.

All-in-all, it was a valuable 2 days, not least for the networking opportunities to meet and kindle relationships with a very diverse range of highly intelligent and highly motivated people.

Looking for some great fiction to read? Get on out there and get yourself a copy of The Bollywood Beauty, written by Shalini Akhil. It’s only just been released, so you can be the first person in your apartment block (or street) to read it! As a measure of its quality, The Bollywood Beauty was shortlisted for the Victorian Premier’s Literary Awards for an Unpublished Manuscript in 2004.

And if that’s not enough for you, Shalini was also the person who beat Peter Malkin in the Raw Comedy finals a few years back . Those of you in Melbourne can also catch Shalini doing some readings at The Age Melbourne Writers’ Festival later this month.

Hooray! JetBrains have finally released IDEA 5 to the public. The new features I’m most looking forward to? Subversion integration, J2ME support (or should that be JME support now?) and of course, the usual smattering of new refactorings, like the ability to safely move non-static methods between classes. Of course, there’s also the well-publicised new JSP and CSS support. Yay!