Sunday, July 08, 2012

'Big Data' - Autodesk take over (and index) the world

Reproduced from my Atkins blog of 27.06.2011
I attended Thursday 24 June 2011’s session for Autodesk’s major account holders, held at their offices in Farnborough. A team from California took Atkins, Arup, WSP, Bam, Laing-O’Rourke, etc. through ‘BIM 360’ for a couple of hours – their thoughts on the way Autodesk’s products will evolve to handle all project data from conception through to completion, handover, maintenance and decommissioning.
Ideas ranged from near-, through mid- to long-term predictions. Some of the key points that leapt out for me were:

  1. Most recent attention has focused on the BIM model and data that can be directly related to it (schedules, specifications, analysis models, etc). Increasingly Autodesk are thinking about the cloud of other stuff that our projects generate – correspondence, photos, calculations…. How will all that become a useful reference for the future?
  2. The session talked a lot about Vault, Autodesk’s offering for organising and sharing all this data within a team. I had not heard of this before. One aim has been to allow team members to feel that they are still working on Windows drives or SharePoint in the same way they have done before. Behind the scenes versioning, audit trails and records are happening – they have almost found it a problem that people have not known they are using Vault (good or bad?).
  3. When sharing data further afield Buzzsaw is their application. It was interesting to hear the Autodesk team talking through the relationships between Revit, Navisworks, Vault, Buzzsaw, etc. They recognise they have had too many products out there but evidently feel they now have a coherent story to tell.
  4. The order of priority when aiming for best practice data management is 1) People, 2) Process and lastly Technology. All too often we focus on the last two first and then wonder why teams don’t eagerly adopt what is given to them. Over half of Autodesk’s income now comes from consultancy and this will rise further - interesting parallels with Microsoft and IBM!
  5. All this talk of structuring data really emphasises the need for us to evolve Information Managers – a kind of twenty-first century Drawing Office Clerk. Since the focus has been on the BIM model to date the CAD manager has been playing v.1 of this role, but this is not going to be appropriate as we begin to talk about broader data relationships. Evidently a new kind of professional may need to emerge.
  6. Anything allowing mobile access to data is currently a ‘killer app’ and these are being hungrily downloaded from Autodesk’s servers by site based teams. In order to remove the need for these apps to download complete files they create location driven ‘bubbles’ of data showing what is relevant to where you are. Future versions will give maintenance teams access to all drawings and specs at their location straight out of the BIM model.
  7. Always nice to pickup the latest language from Silicon Valley. Some parts of Vault have been trialled as ‘Freemium’ versions given away with other software in order gain early feedback. The description of overconfident data managers sitting back and ‘stroking the cat’ had the Americans choking on their mineral water.
  8. Finally cloud based processing now seems to be arriving. I’ve not yet seen anyone sending rendering or analysis off to the cloud but I’m sure I’ll see it within the year. Perhaps it will be a way to beef up to ‘oomph’ of mobile devices such as Ipads way beyond their own hardware specs.