Paul Hillier, ARRB Group, Australia ‘Data Led, Experienced Tuned’ – Developing an Effective Local Skid Resistance Management Strategy 3rd International Road Surface Friction Conference 2011 Paul Hillier, ARRB Group, Australia
Presentation outline Introduction – the challenge faced The value of data (e.g. Organisational Governance Model) ‘Data mining’ Austroads Guide (Jan 2005) – strategy framework Where does data come in? Limitations of data Local experience as a fine tuning tool Closing thoughts
The challenge faced ….. Identifying, collecting & analysing available data Local knowledge and experience
Organisational Governance Model (OGM)
Organisational Governance Model (OGM)
Data mining …… a simplistic view Available Data Facts and Knowledge
Data mining …… a simplistic view Available Data Facts and Knowledge Decisions
Data as our comfort blanket? We are very comfortable as engineers making decisions on data It makes sense……. But what if we can’t afford to collect it? Or fail to recognise that we do not have the full / current picture? How do we integrate practical experience? We’ll come back to this…….
Focus on Skid Resistance Strategies
Austroads Guide – Jan 2005
Austroads Guide – Jan 2005
Austroads 2011 (AT 1131) – data usage Determining zones of skid resistance demand Determining testing regime Setting ILs from first principles (crash rates) Periodic review and update of ILs Identifying and prioritising sites for further investigation
Limitations of data Generic Specific -in managing skid resistance
Generic limitations of data Can include: is not available (or limited availability) ‘ageing’ – network changes have occurred collected by a device out of calibration / maintenance collected by a device that is now superseded data appears (or is obviously) erroneous, skewed or unrepresentative data has processing / collation and/or display errors is too limited to permit patterns and trends to be identified gaps in data are significant data has been previously partially processed or filtered ‘raw’ data can no longer be processed as software etc not available And what if we don’t recognise any shortfalls? Danger of ‘rubbish in, rubbish out’?
Specific limitations of data – Skid Resistance Author’s position - use data wherever possible (and applicable)! But understand / accept limitations: historical rationale (comparing road sections) ‘snap shot in time’ concept crash data: no. of incidents low + v.difficult establishing causation
Importance of local experience as a fine tuning tool Will help ensure that the strategy: is achievable within current resources meets local conditions and challenges will make a difference operationally meets corporate & operational objectives ‘will fly’ (is implemented) is ‘owned’ locally will be continually improved
Closing Thoughts
Author’s experience (skid resistance) Use data in forming strategy and making policy decisions But, understand its possible limitations Ignore local practical (delivery) experience at your peril Involve a number of people – explain the corporate position and requirements, but also understand their local position Get many views - collate them / analyse them Vital for credibility, ownership and achievability Don’t forget the roll out and training regime