Thursday, April 15, 2010

Research Workflow

This morning I have been sitting in ANU postgraduate seminars. These are an opportunity for the students to report what they are doing and get input from other students, their supervisors, but more importantly from people outside their immediate area of research. These seminars can be very stressful for the students, having to deal with everyone from specialists in their specialisation to those who have never heard of the field before.

Questions can be very blunt and negative: "What is your thesis? Who are your supervisors? That is an old text, where are your recent readings?". Some of the comments say more about the questioner than the presenter, such as "I am one of your supervisors, why have I not seen you before today?".

There seems to be far fewer support tools for students doing research than coursework. A student doing a course is given a timetable with deliverables at set dates and an assessment outline. These are now implemented with Learning Management Systems, such as Moodle. It should be possible to use the same tools for researchers. This would set out when the student has to report and would keep a copy of the reports and the supervisor's response. Tools such as Moodle may be too inflexible for research and tools such as Mahara too inflexible.

It may be useful to incorporate some of the features of academic publishing support systems such as OJS. These have a work-flow built in, with editors allocating papers to reviewers, which is similar to the research process. The system will automatically remind reviewers that they are required to submit work (unlike Moodle, which simply records when work is submitted).


Operating robots with virtual reality

Bagus Manuaba is researching at the ANU how to remote control a machine using virtual reality. Devices such as bomb disposal robots are normally operated by remote control, with an operator looking at a live video display and pushing buttons and operating a joystick. The operator is only a few hundred metres from the robot and so there is little delay transmitting video from the robot to the operator and the operators instructions back. However, as the distance in increases the, delay does also.

Where the operator is on one side of the world and the robot the other, the delay prevents fine control and the robot has to be semi-autonomous. In addition, looking at a flat 2d image makes it hard to control a robot.

Some systems use stereoscopic video, such as the Airbus Military Aerial Refuelling Boom System (ARBS) fr the Airbus A330 Multi Role Tanker Transport (MRTT) aircraft being acquired by the Royal Australian Air Force. With this system the operator sits in the aircraft cockpit and fly the boom while looking at a stereoscopic image. Previous systems, such as the KC-135, require the operator to be in the tail of the aircraft. It might be interesting to enhance the Airbus system with virtual reality.

Teleoperation System with Supervisory Control in a Mixed Reality Environment

Bagus Manuaba (SoCS CECS)

CS HDR MONITORING Info & Human Centred Computing Research Group

DATE: 2010-04-15
TIME: 11:30:00 - 12:00:00
LOCATION: Ian Ross Seminar Room

Mixed reality is a new innovation of virtual reality. By combine this innovation with supervisory control in teleoperation system is expected can solve any limitations that occur in teleoperation system today. In this presentation I will give a brief description about teleoperation, supervisory control and mixed reality, and also bring several examples of research that related in this area. In addition, it is also showing the focus and the expected outcomes from my current research.

Labels: , , ,

Mutual Engagement in Mobile Music

Greetings from the ANU in Canberra where Ben Swift from the Information and Human Centred Computing Research Group is talking on "Engagement Relationships in Mobile Music". He is describing the "flow" which musicians experience when performing in a group and are "in the grove". He is researching this experience with multiple iPhones linked by WiFi to a "Viscotheque" server.

The Viscotheqe app for the iPhone allows for manipulating audio sampling. The samples from each iPhone of each performer are mixed by the server for a performance. Three groups of three performers tried the system and data was collected. Video cued recall was used t collect participant responses (video of the participants was played back while they commented to an interviewer). Ben discussed the results in terms of Bryan-Kinns work on enguagement networks. Ben is also involved in the organising of the Australasian Computer Music Conference 2010 at ANU, 24-26 June 2010 in Canberra.

This is an entertaining and scholarly use of ICT. It would make a good project for commercial development through the Innovation ACT program.

The app seems to use only the iPhone's touch screen interface. A useful extension would be to use the motion sensors, for an "Air Guitar" interface. Also it would be interesting to try with musicians blindfolded, or who are blind. Also it might be useful to aurally code each iPhone, so the musicians know who is playing what.

Engagement Relationships in Mobile Music

Ben Swift (SoCS CECS)

CS HDR MONITORING Info & Human Centred Computing Research Group

DATE: 2010-04-15
TIME: 11:00:00 - 11:30:00
LOCATION: Ian Ross Seminar Room

Multi-user interactive music systems offer the possibility of accessible and engaging group experiences. Based on a recent user study of Viscotheque, an iPhone- based system for group musical creativity, this talk will present an analysis of engagement relationships in a group creative task and suggest directions for future research in this area. ...

Labels: , ,

Thursday, April 08, 2010

Greening mutlicore computers

Greetings from the ANU in Canberra where PhD student Xi Yang is giving a seminar on how to use software to save power used by adding a low power cores to a computer architecture. He argues that around 2000 CPU chip designers could no longer increase performance by simply increasing the frequency the chip operated at. Instead designers included multiple cores on the one chip. Normally multiple cores of the same core designs are used. However, a better result might be obtained by using low power cores for routine tasks. The simplest example is a chip with one large core and one small core. The small core then handles routine tasks, such as slow input/output, freeing the big core for handling high performance application processing.

This technique seems to me to have considerable potential. Apart from increasing the performance of a system, it could be used to reduce the energy consumption. I use a netbook with a low power Intel Atom processor. This is more than adequate for routine web browsing, email and word processing, but is not sufficient for large application tasks. Having a high performance core would allow for occasional large tasks.

From a practical point of view a small low power processor is all most desktop computer users need. However, they will buy a higher performance unit just in case they need it. This computer will then waste energy by running idle most of the time. If they could buy a computer with a low power and high power core that would save energy, as the high power core would be unused most of the time.

Some desktop computers and laptops already include an ancillary low power chip, but this is only used to boot the computer using a stripped down operating system (usually a version of Linux or Windows CE) for quick casual browsing. The ancillary chip is unused when the main processor is active. It may be feasible to reprogram one of these systems to provide for both processors working at once. A might be to use several low power cores in place of the high performance core. As an example, eight Intel Atom cores might be provided, but normally only one would be used.

Seminar Details

OS-assisting Core: Improving the Power efficiency of Single ISA Asymmetric Chip

Xi Yang (School of Computer Science, CECS ANU)

CS HDR MONITORING CompSys Research Group

DATE: 2010-04-08
TIME: 10:45:00 - 11:15:00
LOCATION: Ian Ross Seminar Room

The power efficiency of asymmetric multicore systems can be improved by executing suitable operating system workload on low power cores. Devices interrupt handling, IO processing, scheduler and background kernel threads are suitable candidate workload.

Labels: ,

Google Wave Jobs in Sydney

Pamela Fox has mentioned that Google have jobs available in their Google Wave team at their remarkable Google Sydney building for: a Developer Programs Engineer and a Associate Program Manger. My ex-students from ANU would have a head start with these as several ANU graduates already work at Google and the Google team drop in a couple of times a year to give us workshops and seminars.

Labels: , ,

Friday, March 19, 2010

University engagement with industry

Greetings from the Australian National University were the College of Engineering & Computer Science is having an Industry Engagement Day. The idea is to working out how to apply the research done by universities and places like NICTA, to industry. I have some background in this having been involved in formal discipline bodies to change IT research and informa discussions leading to the the creation of NICTA. More recenelty I have helped teaching innovation a ANU.

The first speaker was Professor Mandy Thomas, PVC Research. She pointed out that the Australian Government released an innovation policy "Powering Ideas: an innovation agenda for the 21st century" (12 May 2009) and rearranged portfolios to place research with industry. ANU is looking at ways to link with industry. Support is provided to researchers to seek out industry partners for ARC Linkage Grants. Professor Thomas invited suggestions on how to improve this program.

My suggestion was be to provide training for researches on how to innovate. ANU's engineering students run an excellent Innvoation ACT program each year, to teach how to take an idea to business. This is now an ACT wide program involving other universities and all disciplines. The presentations are recorded and I suggested it would not be difficult to turn this into a formal course.As an e-learning course this could be avialible to all ANU postgraduate students, and students at partner unviersites worldwide.

From the Innovation Report:
... Australia’s recent innovation performance has been uneven, and we have failed to keep pace with the rest of the world. In the last eight years, Australia has slipped from fifth to eighteenth in the World Economic Forum’s Global Competitiveness Index. Our multi-factor productivity grew 1.4 per cent a year on average between 1982–83 and 1995–96. Growth has averaged only 0.9 per cent a year since then, which is no better than we achieved in the 1960s. Since 2003–04, our productivity has actually declined.

... The Australian Government has adopted seven National Innovation Priorities to focus the production, diffusion and application of new knowledge. All of these priorities are equally important. They address the country’s long-term weakness in business innovation, and in collaboration between researchers and industry. The National Innovation Priorities complement Australia’s National Research Priorities, which help focus public-sector research.

Priority 1: Public research funding supports high-quality research that addresses national challenges and opens up new opportunities.

Priority 2: Australia has a strong base of skilled researchers to support the national research effort in both the public and private sectors.

Priority 3: The innovation system fosters industries of the future, securing value from the commercialisation of Australian research and development.

Priority 4: More effective dissemination of new technologies, processes, and ideas increases innovation across the economy, with a particular focus on small and medium-sized enterprises.

Priority 5: The innovation system encourages a culture of collaboration within the research sector and between researchers and industry.

Priority 6: Australian researchers and businesses are involved in more international collaborations on research and development.

Priority 7: The public and community sectors work with others in the innovation system to improve policy development and service delivery.

... the Australian Government will ...


  • Progressively increase the number of research groups performing at world-class levels, as measured by international performance benchmarks.

  • Use mission-based funding compacts and other funding mechanisms to promote collaboration by encouraging universities to organise themselves into research hubs and spokes, and to pursue opportunities to undertake industry-driven research more vigorously.

  • Progressively address the gap in funding for indirect research costs — starting by augmenting the Research Infrastructure Block Grants Scheme with a new Sustainable Research Excellence in Universities Initiative.

  • Help smaller and regional universities develop their research capacity by teaming up with other institutions — supported by a new Collaborative Research Networks Scheme.

  • Increase the capacity of public research organisations, especially to tackle complex problems, participate in domestic and international collaborations, and undertake multidisciplinary research.

  • Continue to invest in research infrastructure to support collaboration and give Australian researchers access to the latest technology, guided by the Strategic Roadmap for Australian Research Infrastructure (2008) — building on $580 million for university research and teaching infrastructure in the first round of the Education Investment Fund, $321 million for research infrastructure in the second round, and $901 million for projects identified through the roadmap and funded under the Super Science Initiative; the third round of the Education Investment Fund will be conducted in 2009–10 to maintain the momentum.


  • Develop a research workforce strategy to address expected shortfalls in the supply of research-qualified people.

  • Increase the stipend for Australian Postgraduate Awards — with an increase of more than 10 per cent announced in the 2009–10 Budget, lifting the stipend to $22,500 in 2010.

  • Significantly increase the number of students completing higher degrees by research over the next decade — building on the Government’s ambition to lift the proportion of 25–34-year olds with a bachelor’s degree and its new incentives to get undergraduates studying maths and science (both of which will enlarge the pool of students qualified to undertake research degrees), as well as its action to double the number of Australian Postgraduate Awards in the 2008–09 Budget.

  • Create viable career paths for Australian researchers — building on the Government’s measures to support research trainees (more Australian Postgraduate Awards with higher stipends), early-career researchers (Super Science Fellowships), mid career researchers (Future Fellowships), and senior researchers (Australian Laureate Fellowships).


  • Introduce mission-based funding compacts that allow universities to determine their own research and collaboration agendas in line with national priorities.

  • Implement Excellence in Research for Australia to measure the quality of university research and guide the allocation of resources.

  • Require universities to provide more meaningful data on research costs through activity-based reporting, and to meet specific performance targets to be developed in consultation with the sector.

Business innovation

... the Australian Government will ...
  • Aim to increase the proportion of businesses engaging in innovation by 25 per cent over the next decade — building on initiatives including Enterprise Connect, Clean Business Australia, and the new $4.5 billion Clean Energy Initiative.

  • Aim to increase the number of businesses investing in R&D over time — fuelled by the introduction of a new R&D Tax Credit, which will double the tax incentive for small-business R&D (restoring it to pre-1996 levels), and lift the base tax incentive for R&D by larger firms.

  • Support innovative responses to climate change — including through Clean Business Australia, the Green Car Innovation Fund, the Clean Energy Initiative, the Global Carbon Capture and Storage Institute, and the Climate Change Action Fund.

  • Improve innovation skills and workplace capabilities, including management and leadership skills — building on Enterprise Connect and the Education Revolution.

  • Support the efforts of Australian firms to get their ideas to market — through initiatives including Climate Ready, the Green Car Innovation Fund, and the new Commonwealth Commercialisation Institute.

  • Work with the private sector to increase the supply of venture capital — building on the Government’s measures to maintain stability and liquidity in the Australian financial system during the global financial crisis, and on the new Innovation Investment Follow-on Fund.

  • Maintain a continuous dialogue with industry about how we can maximise business innovation — including through Enterprise Connect, Industry Innovation Councils, and working groups like that established for pharmaceuticals.

Public sector innovation

... the Australian Government will ...
  • Take advice from the Australian Public Service Management Advisory Committee and the Australian National Audit Office on how the public sector can implement the recommendations of the Review of the National Innovation System.

  • Use public procurement to drive research, innovation and technology development by Australian firms — building on the new Commonwealth Procurement Guidelines released in December 2008.

  • Take steps to develop a more coordinated approach to Commonwealth information management, innovation, and engagement involving the Australian Government Information Management Office and other federal agencies.

  • Consider options for reforming the Australian patent system to increase innovation, investment and trade; and supporting intellectual property education for researchers and business.

  • Improve the management and regulation of biotechnology and nanotechnology to maximise community confidence and community benefits from the use of new technology — starting with a new National Enabling Technologies Strategy.


... the Australian Government will ...
  • Aim to double the level of collaboration between Australian businesses, universities, and publicly-funded research agencies over the next decade — building on initiatives including mission-based funding compacts for universities, Enterprise Connect (including its Researchers in Business Program), Industry Innovation Councils, the new Joint Research Engagement Scheme, and the new Royal Institution of Australia.

  • Increase international collaboration in research by Australian universities — building on actions to open important Australian Research Council awards and fellowships to international applicants, and increase multilateral engagement (for example, in the Square Kilometre Array radio-telescope project).

  • Renew the Cooperative Research Centres Program along the lines proposed in Collaborating to a Purpose — building on the new program guidelines released in 2008, which reinstate public good as a funding criterion, encourage research in the humanities, arts and social sciences, and increase the program’s focus on the needs of end-users.

  • Improve Enterprise Connect’s services to individual firms, anticipating that Enterprise Connect will continue to develop and may include regional clusters and networks uniting businesses, researchers and educational institutions.

  • Promote proven models for linking public and not-for-profit researchers with industry and the wider Australian community — including the CSIRO’s National Research Flagships and the CSIRO ICT Centre.


... the Australia Government will ...
  • Strengthen the Prime Minister’s Science, Engineering and Innovation Council, especially its capacity to look over the horizon and identify emerging trends.

  • Use the Commonwealth, State and Territory Advisory Council on Innovation to improve intergovernmental coordination, starting with the design and delivery of business programs.

  • Give the interagency Coordination Committee for Science and Technology more responsibility and rename it the Coordination Committee on Innovation.

  • Increase the use of metrics, analysis, and evaluation to inform policy development and decision-making.

The future

By 2020, the Australian Government wants a national innovation system in which:

  • the Commonwealth clearly articulates national priorities and aspirations to make the best use of resources, drive change, and provide benchmarks against which to measure success;

  • universities and research organisations attract the best minds to conduct world-class research, fuelling the innovation system with new knowledge and ideas;

  • businesses of all sizes and in all sectors embrace innovation as the pathway to greater competitiveness, supported by government policies that minimise barriers and maximise opportunities for the commercialisation of new ideas and new technologies;

  • governments and community organisations consciously seek to improve policy development and service delivery through innovation; and

  • researchers, businesses and governments work collaboratively to secure value from commercial innovation and to address national and global challenges.

From: Executive Summary, Powering Ideas: an innovation agenda for the 21st century, 12 May 2009
Next speaker was Professor Chris Baker. He started by citing Stanford University's 2006 strategic plan, which while pointing out the unviersity's impressive role in creating new industries and educated captains of industry were working to improve firther. Professor Baker also used the example fo te Cambridge Computer Lab, where about one third of funding come from industry. The comparison with Cambridge is an interesting one. Some years ago on a visit to Cambridge, Sir Peter Swinnerton-Dyer, Chair of the University IT Committee recommended the report "The Cambridge Phenomenon". I found a copy of the report in the ANU library and made a brief study of it:

Segal Quince & Partners (Segal 1985) looked at four policy issues:

  • The role of small new technology based firms
  • Links between industry and higher educational and research institutions
  • The contributions and roles of the public and private sectors in stimulating technological change and economic development. What is the impact of the allocation of allocation of public research funds?
  • The spatial distribution of high technology industry. Will there be a trend away from established industrial and urban areas to attractive rural areas?

Lessons of History

  • There is a long history (100 years) of high technology companies in Cambridge, due to the University.
  • The University is dominant in the city of Cambridge and is strong in scientific fields.
  • The region was already growing before the latest hi-tech developments.
  • Planning which limited large industrialisation may have helped small hi-tech firms.
  • Problems of preservation v development remain.
... two kinds of company links:
  • People forming new start-ups from existing companies, the University, or research laboratories
  • Subsidiaries of existing companies in the area created, but operating essentially as independent companies
... ``nursery units' ... technology parks were:
  • The park was developed in response to demand, not to encourage it,
  • Private sector development was dominant,
  • The buildings were not especially high technology in design or facilities.

Definition of the Phenomenon

  • Large numbers of high technology companies around Cambridge for computer hardware, software scientific instruments, electronics and biotechnology
  • Young, small, independent and indigenous companies
  • Decades of high technology company start ups
  • Research, design and development activities or small volume high value production
  • Links between firms, the university and research organisations...
From: The Cambridge Phenomenon, Summary of The Report, From Net Traveller, Tom Worthington, 1999

Based on this I proposed "Building Arcadia: Emulating Cambridge's High Technology Success", some of which was incorporated in NICTA. Some of the lessons from Cambridge were not welcome in academia and in incorporated. As an example, one reason for the creation of so many start-up companies in Cambridge was the lack of tenure for most staff. Researchers who wanted to keep the Cambridge lifestyle had to go out and set up a compnay in order to earn a living.

Pault Stapleton, from NICTA was the next speaker on the NICTA commercialisation model. NITA has licencsing of technology to existing companies and spinouts (creating new companies). More recently NICTA has offered R&C services. NICTA has a commercialisation team to support the researchers.

NICTA's Investment Model

Grants Pre-Commercialisation

Up to AUD 50K Market Development Grant Up to AUD 100K Proof of Concept
Investments for Commercialisation
Up to AUD 250K NIPR Investment
Up to further AUD 250K from NIPR

From: "NICTA's Investment Model", NICTA, 2008.
As an example Paul mentioned the Goanna Software Bug Detection tool , AutoMap map analysis, and the Performance Assessment for Service Architecture (ePASA) for ICT development.

Labels: , ,

Wednesday, March 03, 2010

Academic papers and citations

One of the tertiary institutions I teach at is applying for accreditation, so I have to submit a CV. The CV template asks for a publication and citation counts. While I have written a couple of books and dozens of articles, I have never worried if these count for academic purposes. I am not a career academic (only an adjunct) and receive no payment for research publications. So I entered "nil" on the form.

That looked a bit poor, so I thought I should look to see if I had written some papers which counted or had been cited by some formal papers (given that thouadns of people read my web site each day I assume someone must mention it somewhere). A search of Google Scholar listed one of my books:

[BOOK] Green Technology Strategies: Using Computers and Telecommunications … [HTML]
T Worthington - 2009 -
This book is about how to reduce carbon emissions and achieve other environmental benefits
by using computers and telecommunications technology. It is designed to be used within an
online course for profes- sionals, using mentored and collaborative learning techniques. ...
There were also 24 citations of my web pages in schollarly web pages:
Unfortunately none of these appear to count for academic purposes.

Eventually someone suggested the "H-index" which lead me to a reference to "Web of Knowledge". A search of that shows one paper:
Source: AUSTRALIAN COMPUTER JOURNAL Volume: 27 Issue: 1 Pages: 16-16 Published: FEB 1995

Times Cited: 0

Presumably none of the other publications which cited me were formal papers and so do not count. But one paper is better than none.

Labels: ,

Saturday, January 30, 2010

Stack iPads for Library e-Book Display

Apple iPad video wall proposed by Clarke Hopkins ClarkeAccording to Gizmodo, architects Clarke Hopkins Clarke have suggested stacking hundreds of Apple iPads on a library wall to display iBooks. A wall of iPads in a library is an intriguing idea, but would cost twenty times as much as some alternatives. There are much more affordable and environmentally efficient ways to build a video wall to display e-Books.

Many libraries now use large flat panel displays. These can be used to show book details. The latest of these displays used LED backlit LCD panels, which use less power than old plasma screens. One computer can drive many displays, making the setup much cheaper. A video projector can also be used to make a wall size display which can show one large image, or well as many small ones. At the Australian National University's famous " CSIT building N101 seminar room" I have used the full wall display for presentations. Three high resolution projectors cover one wall of the room and a computer with three video interfaces knits these into one large desktop. The wall has also been used for video art display. The wall can display cinema style video, or when the room is needed for other purposes, simply switched off.

An Apple iPad has an area of about 0.05m2 and costs about US$500. The Dell G2410 24-inch LED LCD monitor has an area of 0.23m2 and costs about US$300. Allowing for the extra computer hardware to drive a video wall, the LCD screens would cost about US$500 each. These would cost the same as the iPads, but because the iPads are much smaller the wall would cost five times as much.

The wall depicted by Clarke Hopkins Clarke has thirty columns of iPads, ten high, or 300 in total. The wall would cover 15m2 and the iPads would cost US$150,000. The same wall area would require about 66 LCD screens and cost US$33,000. Using video projectors would cost about US$7,000.

Labels: , , , , ,

Tuesday, January 26, 2010

BarCamp Canberra 2010 on e-Gov 2.0

A BarCamp Canberra 2010 around the topic of e-Government and Web 2.0, will be Saturday, 6 February 2010 from 9:00 am in the famous Room N101 of the School of Computer Science, Australian National University. This a free event where anyone can turn up and offer to speak. I attended Bar Camp Canberra 2 last year and this year have volunteered to speak on:
e-Books for e-Learning

Tom Worthington shows how he used simple web pages and free open source software to create a university level e-learning course and accompanying e-book for the Amazon Kindle, Google Android, Apple iPhone, i-Slate and Netbooks.

Labels: , , , , ,

Monday, December 14, 2009

Supercomputer from game console components

Greetings from the famous room N101 at ANU where Wayne Luk from Imperial College London is talking on "A Heterogeneous Cluster with FPGAs and GPUs". He started by apologising the talk will not be polished as the work is very new and they are just starting to get results. He then gave us a quick tourist's guide to Imperial, which is near Kensington Palace and the Albert Hall. He argues that techniques for embedded systems could be applied to high performance computing. This is counter-intuitive as embedded computing is usually used for low cost small scale computing in consumer goods, whereas supercomputers have been made from high cost, high performance custom components.

The concept is that an application written in a conventional programming language would be compiled partly into code for a conventional processor and partly into configuration information for customisable chips. This could be used for applications from supercomputers to distributed applications using "smart dust".

The application would used Field-programmable gate arrays (FPGA). These are now used in consumer equipment, such as LCD TVs. FPGAs are very efficient in terms of cost and processing power per unit of energy used. But programming FPGAs is complex. FPGAs have high speed serial interfaces which allow them to be used together. Examples are the Stratix III and Stratix IV. Imperial have produced an 8 x 8 "cube" of FPGAs for emulating processors ("MUMAlink" Interconnect Fabric), and for prototyping the entertainment system in a car.

raphics processing units (GPUs) have multiple processors, a shared bus and memory on a chip. As a result they are less customisable and less power efficient than FPGAs, but they are easier to program. Ideally FPGAs and GPUs would be combined with conventional processors in the one system for maximum flexibility. This approach differs to the one investigated in "Comparison of GPU and FPGA hardware for HWIL scene generation and image processing" (by Eales and Swierkowski, DSTO Weapons Systems Division, 2009).

Imperial has a 16 node cluster "Axel", with an AMD CPU, C1060 GPU and Vpf5 FPGA, connected by Gigabit Ethernet and Infiniband on the FPGA. This has a "non-uniform node" architecture: there is a CPU, GPU and a FPGA in each node, with these connected on a common backbone. Initially a Single Program Multile Data design was used for simplicity. The backbone has Gigabit Ethernet plus Infiniband.

Linux runs on each node, using NFS. There is a custom resource manager and public domain cluster manager (openMP and OpenMPI). There is a communications bottleneck with data having to pass through the CPU from the FPGA to the GPU. Direct communication would be desirable but difficult.

The question then is what are common patterns of parallelism which the system could support. The "Berkeley Dwarfs" offers a set of common patterns.

The new Intel Atom chip (codenamed "Pineview") due in early 2010, is rumoured to have an integrated graphics core, which could be useful for low cost systems.

Iridium is planning a new generation of communication satellites with provision for an earth observation payload. It might be interesting to see how much processing could be usefully put on-board. The processing might be reprogrammable to to communications or processing as required and depending on where in the orbit they are. The Iridium satellites can only carry out their primary function of communications during a small part of their orbit. The rest of the time the satellite could carry out observations and process data.

Labels: , ,

Sunday, December 13, 2009

Building a supercomputer from game console components

Wayne Luk from Imperial College London will talk on "A Heterogeneous Cluster with FPGAs and GPUs" at the ANU, 14 December 2009. GPUs (Graphics Processing Units) are used to offload complex image processing from the main processor in PCs and games consoles. FPGAs (Field Programmable Gate Arrays) are more flexible devices which can be reconfigured for custom applicators. These cips have become popular as a way to design low cost specalised supercomputers. But no one is exactly sure of the best design for such a supercomputer. Thus the need for research to find out how. Apart from research these systems have applicaiton in predicting climate change and cracking encryption codes.
                   Seminar Announcement
School of Computer Science, CECS
The Australian National University

Date: Monday, December 14, 2009
Time: 11:00 am to 12:00 noon
Venue: Room R214, Ian Ross Building [31]

Speaker: Wayne Luk

Title: A Heterogeneous Cluster with FPGAs and GPUs


This talk describes a heterogeneous computer cluster called Axel. Axel contains a collection of nodes; each node can include multiple types of accelerators such as FPGAs (Field Programmable Gate Arrays) and GPUs (Graphics Processing Units). A Map-Reduce framework for the Axel cluster is presented which exploits spatial and temporal locality through different types of processing elements and communication channels. The Axel system enables experiments involving FPGAs, GPUs and CPUs running collaboratively for applications in high-performance computing, such as N-body simulation.


Wayne Luk is Professor of Computer Engineering at Imperial College London. He was a Visiting Professor at Stanford University. His research interests include theory and practice of customizing hardware and software for specific application domains, such as multimedia, financial simulation, and biomedical computing. He is a fellow of the IEEE and the BCS.

From: "A Heterogeneous Cluster with FPGAs and GPUs", ANU, 2009

Labels: , ,

Wednesday, December 09, 2009

What we want web-wise

Tim Jones will talk on "Using Firefox and JavaScript to Combat Apathy among Experiment Participants" at the ANU, 10 December 2009. This work has wide applicability to anyone wanting to find out what people like to look at on the web. I recommend the talk for government and commercial web practitioners, as well as academics and researchers.

It is good to see someone is worrying about real research, while the rest of us are at the government broadband talk-fest in Sydney. ;-)

Seminar Announcement

School of Computer Science, CECS
The Australian National University

Date: Thursday, 10 December 2009
Time: 4:00 pm to 5:00 pm
Venue: Room N101, CSIT Building [108]

Speaker: Tim Jones

Title: Using Firefox and JavaScript to Combat Apathy among
Experiment Participants

A simple way to conduct web search quality experiments is
to log user behaviour when using a new front end (that we
provide) to a commercial web search engine. Unfortunately,
users tend to stop using these new front ends after the
initial novelty period has worn off. This makes it difficult
to collect data about real user behaviour when using web
search engines.

In this talk, we firstly describe the widely used experimental
framework that users seem to stop using. Next, we investigate
this drop off in user participation, discuss some probable
reasons, and then present a couple of ways one might alter
the experiment design to ensure users continue to participate.
We then illustrate the design and construction of a Firefox
addon to augment search engine result pages. The addon is
written in JavaScript using the popular jQuery library, which
makes data acquisition very simple. The talk will include
a quick jQuery primer, as it is a useful tool for anyone
wanting to manipulate or extract information from web pages.

Timothy Jones is a PhD student in the ANU Department of
Computer Science. His primary research interest is Information
Retrieval, but he also enjoys a piece of cake and a good cup
of tea. ...

From: Using Firefox and JavaScript to Combat Apathy among Experiment Participants, ANU School of Computer Science, 2009

Labels: , ,

Friday, December 04, 2009

Propose an e-Oxbridge education

Having spent some of the week discussing the future of higher education, with Professor Paul Ramsden and my colleagues at the Australian National University (including my contribution on "Forums and Feedback for e-Learning"), I felt it was time to suggest a way forward. I have proposed an e-Oxbridge educational model for the ANU School of Computer Science (SoCS) .

SoCS has ambitious goals set for "unique", "advanced", "interdisciplinary" and "research lead" undergraduate and masters courses. To achieve this, I have proposed a computer enhanced version of the "Oxbridge" model of education. With this approach at Oxford and Cambridge Universities (where I have given the occasional seminar) students are part of a community of scholars, write and discuss material with their peers and their tutors each week. This can be adapted to the 21st century:
  1. Human communication: I suggest teaching all students how to research, write and report. While most undergraduates will not go on to postgraduate research and therefore not need to write a scientific paper, they will have to write technical and business reports which require similar skills. Therefore I suggest teaching how to write and present an argument in the introduction to undergraduate and postgraduate programs. I have done some of this in Green ICT, where I get the students to research and discuss issues online and write a reports about a real problem.
  2. Self motivated work: In each course I suggest setting the students a task, giving them the tools and then helping them with the work. In practice this would be done by providing learning materials in traditional written form, as well as multimedia, as used by the "Hubs and Spokes" project. This would then free up staff time to work with the students in small groups and individually. This would also force a discipline on staff, who would need to carefully design course materials in advance. Also this would allow administration to be greatly simplified, with less need for timetabling of classes and resources. This would aid social inclusion, with full and part time students could in the same class, along with domestic, international and remote e-learning students.
  3. Interdisciplinary skills: I suggest designing SoCS programs to fit in with ANU wide programs and those of partner universities. In this way students will be able to study subjects outside Computer Science in other parts of the university.
Instead of developing whole, self contained undergraduate and
postgraduate programs which are exclusive to SoCS, I suggest SoCS have modules which can fit with other disciplines and can be used by other disciplines. A student should be able to do a standard undergraduate or postgraduate program at the ANU which incorporates SoCS education. While the SoCS programmes might have fancy names, such as Bachelor/Masters of Advanced Interdisciplinary Computing", they should underneath be made of ANU standard components. Ideally the courses should be able to be tailored by the students themselves, as is done with ANU Graduate Studies Select.

Labels: , , , , ,

Thursday, November 26, 2009

Low cost computer security device to replace passwords

Greetings from the famous room N101 at the School of Computer Science at ANU. Bob Edwards is presenting on "Yubikey Authentication in a Mid-sized Organisation". This is a preview of paper for Linux Conference of Australia 2010 (LCA2010) in January.

The Yubikey is a low cost ($10) security token designed to replace passwords for computer access. It is a small USB unit designed to be attached to a key ring and inserted into a computer when access is needed. The device generates a 44 character pseudo random number when a button on the unit is pressed. It emulates a keyboard to send the number to an application. The device uses AES-128 bit encryption.

One use which has been poposed is Yubikey identifying airline pilots on the U.S. Department of Homeland Security Transportation Security Administration blog.

Yubikey provide an online authentication server, which can be used. However, as Bob points out, this requires you to trust the security and reliability of Yubikey's system. Yubico allow for the device to be reprogrammed with a new 128 bit key so that an organisation can run its own authentication server.

One limitation of the device is that it has no internal battery and so cannot keep track of time. As a result the tokens generated never expire. Also as with any token, it must be kept physically secure. If left in a computer (as happens with sensitive devices), it will provide access for the next person who happens along (although an additional user entered id and password could be used).

Other limitations of the device are that it requires a USB port. Allowing USB devices to emulate a keyboard creates a security problem, but if disabled would stop the Yubikey working. Also, because it emulates a keyboard, any application on the host computer can read random numbers generated by the Yubikey.

One option which might be interesting for Yubico to make a credit card sized Yubikey. This would have sufficient space to overprint as identity card and have a conventional magnetic stripe. This could then be used with existing standard identity card printers and magnetic stripe security systems for student and staff ID cards. There are a number of designs available which have a USB interface on the edge of the card, or in a flexible cutout or with a folding card. These may seem cumbersome and subject to failure, but I have had a SanDisk Ultra SD Card for some years, which folds in the middle to covert to a USB drive.

Yubico might like to market the devices for green ICT power saving. The host computers could be programmed to switch to low power mode unit the device is inserted. It could also be used with thin client devices where the user's application run on a server. When the device was inserted in a different client, the applications would be restored as they were when suspended.
The Swedish company Yubico manufacture the Yubikey One-Time-Password (OTP) USB device and have released all protocol and other relevant details which makes the Yubikey particularly attractive as a low-cost and non-vendor-lock-in authentication solution.

Bob will demonstrate the Yubikey for the purposes of secure authentication on untrusted end-user systems (eg. PCs at an Internet Cafe or a friends house etc.) and will discuss some of the advantages as well as some of the weaknesses of the Yubikey system. He will then go on to describe the development of an authentication server written in C and based on a PostgreSQL database and implementing LDAP and other authentication protocols. This will include some technical details of how to use the APIs for connection to the database, parsing the ASN1 LDAP queries, dealing with denial-of-service attacks etc. He will also discuss some of the code he has written to implement the Yubikey protocol on devices with no USB port (eg. a PDA or mobile phone etc.).

This talk is a prelude to a paper Bob will present at the Linux Conference of Australia in 2010 (LCA2010) in Wellington, NZ in January.

Bob Edwards is the Chief IT Officer in the School of Computer Science at the ANU. He also teaches into the Computer Networks course and the Free and Open Source Software Development (FOSSD) course, amongst others.

From: Yubikey Authentication in a Mid-sized Organisation, ANU 2009

Labels: , ,

Tuesday, November 10, 2009

Computer art at the Australian National University

Works on The Web and works on The Wall Diverse works by diverse studentsThe official opening of "Works on the Web and Works on the Wall: Diverse Works by Diverse Students" will be by Professor Elizabeth Deane, 6pm, 13 November 2009 in Room N101, CSIT Building, the Australian National University, Canberra. This computer art show is part of the activities of a second year course, Automating Tools for New Media, by Tim Brook, which received an ANU Teaching Enhancement Grant for 2009. Please feel welcome to roll up and enjoy some of the artistic creations of the students.

Labels: ,

Monday, October 26, 2009

SPICE Learning Management System

Greeting from the ANU Software Engineering Showcase. Teams of third and fourth year students are showing off their projects. One of these is the SPICE Learning Management System. This uses Moodle and LAMS to provide a system to assist volunteer teachers in Canberra. The most interesting part of this is that normally the students write software. In this case they decided the original idea of custom software on a bootable flash drive was not a good one and configured an Internet connected web based learning manageemnt system instead. This was a good decision. I have been invvled in sevceral court cases as an expert witness where a system development team did not stop, question what they were doing and choose another path.

One aspect the students do not appear to have realised is that they can still provide paper based materials from the Moodle system.

Labels: , ,

Thursday, October 22, 2009

GovHack: Mashup Australian Government Data on 30 October

GovHack is a free, intensive event at in Canberra on 30 to 31 October. This is sponsored by the Government 2.0 Taskforce and supported by CSIRO. Government agencies, locals, state, federal and international, might like to come forward with APIs and datasets to be used by the participants. Offer datasets and tools via the Wiki. University and industry researchers can contact Laurent Lefort at CSIRO. The event will explore some of the ideas for the use of government information discussed at Bar Camp Canberra.

Labels: , , , , ,

Tuesday, October 20, 2009

Automated Quality Assessment of Obesity Websites

Ramesh Sankaranarayana will give a seminar on "Automated Quality Assessment of Obesity Websites", 4pm, 29 October 2009 at the ANU in Canberra. This follows previous work on "Automated quality rating of depression websites".

ABSTRACT: We previously developed an automated quality rating technique (AQA) for depression websites and showed that it correlated 0.85 with expert ratings using evidence-based guidelines. AQA scores the pages within a site against complex learned 'quality' and 'relevance' queries using the BM25 ranking function. It then aggregates these scores across sites using a learned combining function. In this talk, we report on our generalisation of the AQA method and our evaluation of its application to a different health domain, namely obesity. We find that correlations as high as 0.80 averaged across ten folds can be obtained, but that performance is quite dependent on the choice of the high quality websites used for generating the 'quality' query. This is ongoing work ....

From: Automated Quality Assessment of Obesity Websites, School of Computer Sciecne, ANU, 2009

Labels: , ,

Thursday, October 15, 2009

Concurrent and Distributed Computing at Google

Daniel Nadasi, Google Software Engineer and ANU graduate, will give an open lecture in Canberra this Friday on "Concurrent and Distributed Computing at Google". This is for students of Concurrent and Distributed Systems (COMP2310/6310), but anyone is welcome to attend.


Concurrent and Distributed Computing at Google

Daniel Nadasi (Google)

DATE: 2009-10-16
TIME: 14:00:00 - 15:00:00
LOCATION: Chemistry Lecture Theatre T1

Have you ever wondered how Google's massive computing infrastructure works? How Google manages to keep petabytes of data and thousands of machines stable, consistent, responsive, scalable and fault-tolerant? In this talk, you will be presented with a bottom-up tour of Google. Starting from the building blocks of concurrent and distributed computing, you will see how Google's systems are constructed, through the internal systems that are used, to some of the newest products, such as Google App Engine.

Daniel Nadasi joined Google as a Software Engineer in November 2007. He currently works on advertising for Google Maps. Aside from this, he has also worked on Google Tasks and open source initiatives. Daniel graduated from ANU in 2007 with a Bachelor of Science with a double major in Mathematics."

Labels: ,

Wednesday, October 14, 2009

Software Engineering Showcase

The School of Computer Science, ANU, will hold a
Software Engineering Showcase, 26 October 2009
12.00 - 2.00pm. This will display some solutions to software engineering problems.

Software Engineering Showcase
Monday 26 October 2009
12.00 - 2.00pm
The Australian National University
Ian Ross Building 31, North Road Map Ref
Seminar Room (R214)
Includes refreshments from 1.00pm

You and your colleagues are invited to attend the Software Engineering Students Showcase hosted by the School of Computer Science, ANU.

Our students collaborated with several Canberra businesses to find creative solutions to some of the software engineering problems that the public and private sectors encounter.

The event is an excellent opportunity to meet our students and staff, and for you and your colleagues to find out more about our software engineering degree, and how the theory students learn is being put into practice in the real world.

Refreshments will be served.
RSVPs are essential:
by Friday 23 October 2009.



12.15pm Welcome

12.20-1.00pm Student Presentations

1.00pm Refreshments and tour of projects

1.45pm Close

Labels: ,

Monday, August 17, 2009

Environmental Assessment of a IT Intensive Building

You are welcome to attend this seminar on "An Environmental Assessment of the SoCS at CSIT North". Several Green ICT Sustainability students are doing similar assessments of other buildings on the ANU Campus. The results will then be combined to provide an overall report of the campus. This will likely be the most detailed such student ever undertaken of a university campus:


An Environmental Assessment of the SoCS at CSIT North

Gong Lu and Soraya Nour (ANU)

DATE: 2009-08-21
TIME: 16:00:00 - 16:30:00
LOCATION: CSIT Seminar Room, N101

Part of ANUgreen's Sustainability @ Work Program is to conduct environmental audits of workplaces around the campus. Over June and July 2009, the two speakers conducted an audit of the School of Computer Science at the CSIT North Building. In this seminar, they will report on the results of their audit. After describing the methodology for the audit, they will detail their findings in the main areas of the audit: energy & IT, water, transport, waste, purchasing & paper, and pollution prevention. In each area, they will give recommendations for improving the sustainability of SoCS at CSIT North

Gong Lu and Soraya Nour are volunteer undergraduates at the ANU who are participating in the Sustainability @ Work program.

Labels: , ,

Wednesday, May 27, 2009

ANU Green ICT Seminar Series 2009

This is to invite nominations for presentations in the ANU Green ICT Seminar Series. A six month program of weekly seminars is to be held on Wednesdays 2 to 3pm at the Australian National University in Canberra, as part of the Masters course "Green ICT Strategies" (COMP7310) . The weekly topics for the seminars have been chosen to match what the students are learning that week. Members of the ICT and related professions are invited to nominate to give a presentation, or simply to attend and discuss the issues with the students. There is no charge for attendance at the seminars, however seating priority will be given to the students enrolled in the course.

22 July: Understanding climate science

Understand environmental, social and business context for sustainability, and overview of background, boundaries.

29 July: The Global ICT Footprint

Estimate the embodied carbon and the footprint from use of telecommunications, data centres and desktop PCs.

5 August: Enabling ICT

Investigate how ICT systems can reduce energy and materials use by improving the efficiency of business systems by replacing the movement of goods with information (dematerialisation), improve the efficiency of machines (smart motor systems), logistics, buildings and grids.

12 August: Energy saving - Data Centres and Client Equipment

Computers and telecommunications equipment contributes about 2% to greenhouse gas emissions. Look at how data centres and client equipment can be made more efficient.

19 August: Materials Use

Energy reduction is only part of making a Green ICT system, there is also the issue of use of materials and hazardous substances.

26 August: Methods and tools

Ensure that appropriate methods and tools for the planning, development, operation, management and maintenance of systems are adopted and used effectively throughout the organisation.

2 September: Business process improvement

Recommend alternative solutions which reduce environmental impact, assesses feasibility, and recommends new approaches. ICT has the potential to provide significant environmental improvements, by replacing energy and materials consuming processes with more efficient ICT ones. How do you analyse business processes to identify alternative solutions which reduce environmental impact, assesses feasibility, and recommends new approaches?

9 September: Improving Data Centre Energy Efficiency

Investigate how to implement and assess data centre efficiency.

16 September: Enterprise Architecture

The business of business is business, so any environmental goals have to fit into the systems capability strategy which meets the strategic requirements of the business. How do you incorporate Green ICT into the models and plans to drive forward the strategy, taking advantage of opportunities to improve business performance, as well as environmental benefits?

23 September: Procurement

Much of the environmental benefits come about by selecting the right products and services. How do you write ICT requirement documents to ensure that your hardware, software and services suppliers provide green products?

14 October: Energy Star Program and Quality Management

Ensure that the processes for producing a product or service sustainability will do so to a consistent standard. Investigate the fit between US EPA's Energy Star Program and those of Quality Management Systems.

21 October: Compliance audit

Assess the conformity of corporate systems to environmental standards, such as ISO 14000 series of environment management standards, Electronic Product Environmental Assessment Tool (EPEAT), and Energy Star Program.

Labels: , , , , ,

Tuesday, May 12, 2009

Digital Library and Digital Education in China

Professor Zhendong Niu Greeting from the School of Computer Science at the ANU in Canberra, where Professor Zhendong Niu is visiting from Beijing Institute of Technology. One of his areas of interest is Digital Library and Digital Education. SOme projects of interest are the China National Knowledge Infrastructure, National Cultural Information Resource Sharing Project, China Academic Library Information System, National Foundation Education Grid (for k12 education resources for schools). The Professor also has an interest in legal portals.

What I found of most interest was work on integrating digital libraries with education. That may sound an obvious combination, but not much has been done in this area. The builders of document repositories work very separately from those of learning management systems. One aspect which doesn't arise as an issue is language; the software tools can work in Chinese and English, using the same ontology. An example is the Olympic Games (which BIT was involved in software for), the same concepts apply, even where different words are used in English and Chinese (I suggested using pictograms for the Beijing Olympics).

It was interesting to see the similarities with the issues of technology for education for China with the "Supermarket of E-learning" by China TV and India's use of satellite TV.
With the development of digital library, social networks, and user-generated content, need for trust and reputation models become prime. In this paper, we propose a user reputation model. As an encouraging and sanctioning mechanism, it has been applied to the DLDE (Digital Library and Digital Education) Learning 2.0 Community that is developed by our lab based on digital repositories management etc. The model combines user's individual activity analysis approach and collaborative activity analysis approach. Individual activity analysis approach is used to analyze the activities in which users participate individually and give its evaluation method. Collaborative activity analysis approach is used to analyze users' collaborative activities; three different categories of users' collaborative activities and corresponding evaluation methods were proposed in this paper. Experiments show that the proposed reputation model can accomplish the mission of encouraging good behaviors and differentiating the ability of students. Therefore it can fit well in our Community. ...
From: A User Reputation Model for Digital Library and Digital Education DLDE Learning 2.0 Community, Jin, F., Niu, Z., Zhang, Q., Lang, H., and Qin, K. 2008. , In Proceedings of the 11th international Conference on Asian Digital Libraries: Universal and Ubiquitous Access To information (Bali, Indonesia, December 02 - 05, 2008). G. Buchanan, M. Masoodian, and S. J. Cunningham, Eds. Lecture Notes In Computer Science, vol. 5362. Springer-Verlag, Berlin, Heidelberg, 61-70. DOI=

Labels: , , , , ,

Monday, May 11, 2009

Distance education for millions of unviersity students

Greetings from the famous room N101 at the Australian National University in Canberra, where Professor Uma Kanjilal, Director of the School of Social Sciences, Indira Gandhi Open University (IGNOU) is speaking on "Standards and distance education". The talk is being recorded and I will provide a link here, when available. The Professor is visiting Australia and will speak to federal government agencies later in the day and at University of Canberra, later in the week. The visit is funded by the Australia-India Council, University of Canberra and Flinders University and the purpose is for knowledge exchange around ICT in education. I will post some comments here as the talk goes along.

Initial distance education in India was printed material by post. Integrated multimedia is being added, with TV and radio. It needs to be kept in mind that infrastructure is needed, including libraries and A/V facilities, computers, and Internet.

India has a Distance Education Council to oversee provision of courses. This sets standards for materials, registration processes, support services for learners, ICT infrastructure and assessment. Before creating a course, the institution has to do need assessment to show there is a requirement for a course and who the target group are.

The Indian Distance Education Guidelines are available online:
  1. Norms and Standards for Management Programmes
  2. Norms and Standards for IT Education
  3. DEC-GUIDELINES, for regulating the Establishment and Operation of Open and Distance Learning (ODL) Institutions in India
  4. Norms for ODL
  5. Norms for Online Programmes
The India PM has set up a National Knowledge Commission as a high-level advisory body. There will be an advisory group on Pedological Content and a Technical Advisory Group.

India uses its own satellites for broadcasting educational material. More interactive and feedback techniques as resources permit. Few Indian students have Internet access at home, so some methods being used are cyber cafes. Mobile phone SMS is being used for student support and 3G smart phone support will be offered when these phones are more widely available.

ps: For my own experiences of Indian cyber cafes and wireless networking, see: Living in an Indian Village in Goa for Three Weeks

pps: Unfortunately the battery went flat in my MP3 recorder, so only the first 10 minutes of the talk were recorded: Uma Kanjilal ANU 2009 05 11 (MP3 4.48Mbytes)

Labels: , , ,