Personal tools
You are here: Home Local Information A guide to successful grant applications Report from the "Towards Intelligent Information Infrastructure" workshop

Report from the "Towards Intelligent Information Infrastructure" workshop

A short trip report from James Cheney about the "Towards Intelligent Information Infrastructure" workshop organized by EPSRC (Dec. 3-4), in case it is useful for others who think they might be interested in the topic and associated EPSRC funding opportunities. Alan Bundy and Rob Baxter (EPCC) also attended from Edinburgh, and might have additional observations.

Monday afternoon 

The first day started with an introduction to EPSRC's various shaping/molding initiatives from Liam Blackwell, aimed at early-career researchers who EPRC guesses might not have heard of this.  The rest of the afternoon consisted of presentations and panel-style discussions with the speakers.

The speakers/titles (and my notes about what they talked about) were:

Intelligent Users need an Intelligent Infoformation Infrastructure: Tom Rodden, University of Nottingham

  • need for mature, mundane, predictable platforms
  • abstraction needed to build layers of infrastructure with separation of concerns, but then get in the way of performance
  • hard to understand failures at application layer, layers can't cross-optimize
  • need for collaboration among different fields within CS/ICT - how to promote this?

New Frontiers in Wireless Research: Tim O’Farrell, University of Sheffield

  • lots about bring network infrastructure down to very small cells, using spectrum in new ways, software-defined networks

Scalable Transformation of a Mass of Data into a Proactive Source of Intelligence: Simon Maskell, QinetiQ

  • need for data analysis tools, answering open-ended questions

The era of SDx – the implications of software defined infrastructure: Simon Fletcher, NEC Telecom Modus

  • software-defined networks, 
  • using machine learning, trust, collective intelligence, UI
  • Network virtualization - SDF/OpenFlow (Stanford)
  • Compression techniques, XML, headers in wireless protocols

Reducing the Energy Consumption of the Internet: Richard Penty, University of Cambridge

  • Energy consumption of internet ~ 2% expected to grow to 4%, comparable to air travel
  • Mostly driven by video and wireless 
  • GreenTouch consortium

Computing and Communications On, In and Around the Body: Chris Baber, The University of Birmingham

  • Wearable computing, body-area networks
  • Motor / neuro Rehabilitation - helping people remember how to make cup of tea

As the titles suggest, 4 out of the 6 were pretty hardcore networks people; Rodden was more of a socio-technical systems person working towards the networks end of the spectrum and Maskell was the only one talking about data or analysis (from a intelligence/security perspective). 

Tuesday

Before the workshop, EPSRC asked participants to group a set of topics (presumably identified by them from our applications) using an online tool.  They then used this to somehow generate six groups, and at the end of the first day we each voted for up to three groups, and were allocated to these on the second day.

The groups were roughly:

  • Big data (two of these)
  • Intelligent/advanced networks
  • Network infrastructure/hardware
  • Privacy/security/provenance
  • HCI/socio-technical
  • Software modeling/theory - I believe not enough people voted for this one for it to be allocated as a group (!)

Wednesday

Wednesday: was split into four sessions in breakout groups, with two plenary sessions to present results.  I was in the "Provenance/security" group.

  • Session 1: identifying activity already going on in various areas
  • Session 2: thinking 5-15 years ahead, what should the TI3 priority achieve?
  • Session 3: identifying challenges and obstacles to progress in the various areas
  • Session 4: presenting the challenges & obstacles, and finally voting on the different suggestions outlined by the different groups

Broad themes (topics that seemed to come up from several directions, though this is somewhat biased towards my group's discussion):

  • Need to break down barriers (or open op abstractions) enabling truly cross-ICT-discipline work, supporting government, academic, industrial collaboration
  • Moving intelligence into the network (but how is this different from "active networks" 10 years ago, other efforts that failed because the resulting systems became too complex/unmanageable?)
  • Difficulty of establishing realistic testbeds or models for evaluating socio-technical systems / information infrastructure
  • Need for incentives and tools, curation, provenance, scientific data lifecycle

Stray observations

Interestingly (!) the audience was I think about 90% male.  Not sure what, if anything, this reflects about EPSRC's selection process or advertising about the event…

My impression is that the TI3 call in the spring is likely to be about communications/networking, since that seems to be what the EPSRC people involved in TI3 seem to be most familiar so far; there may be other calls later, focusing on other centers of mass identified in this workshop.  So, making friends with people working on networks now seems wise.  (For example, I think Peter & Mahesh's Tegola project is actually in a good position if things unfold as I suspect; people here working on performance modeling, database techniques for declarative networking, or PL for software-defined networking might also want to keep alert.)

Document Actions