Making a Safety Culture Chart

[batman1056]

I have been given a challenge to develop a chart of some kind that shows the level of safety culture/risk level based upon postive and negative indicators. The challange is to have something that reflects against negative elements (e.g. RIDDOR, Lost Time accidents, Non Lost Time Accidents) and positive proactive elements (e.g. Safety Tours, Inspections, Safety Audits).

Each of the positive and negative elements would have a weighitng e.g. RIDDOR would have a -ive value e.g. -10 and a safety tour would be a +itve value e.g. +1

I am looking for examples or ideas on how to represent safety as a single value. The resulting table should allow top managment to understand the higher the negative value of the total the more probable a major incident would occur.

I did create a speed dial that moved depending on the values - but this was too difficult to normalise. Hence the question. We already have seperate indicators and charts for each of the seperate KPIs (e.g. no of accidents) but I would like to develp a holistic overview sliding chart or something --- any help appreciated.

 

[Steve Prevette]

Normalization / Creation of Indexes is indeed difficult! Why not avoid all that and go to an SPC-based Balanced Scorecard? Keep the leading and lagging indicators grouped together and get a big picture overview. In the midst of doing that at Savannah River and also Chalk River nuclear sites. Employee surveys can also be trended by SPC and included as a leading indicator. See http://www.wmsym.org/archives/2007/pdfs/7148.pdf

 

[batman1056]

Thanks - we have a BSC system in place too with the monthly report showing the clean split between leading and lagging. The last HS Manager put in the safety plan that we would produce a single chart showing the current level of safety and the level of risk probability - he then left without defining out how this would be achieved. I have managed to develop something which sort of works – I will keep trying to figure out the short-term and long term weighting elements. The ideal situation was to have -1 as being the worst case and +1 being we are no danger of anything happening (major) so the value produced would be something in the middle e.g -0.0234

 

[Steve Prevette]

Perhaps there is a reason the last safety manager left. Very easy to TALK about setting up a single metric, very hard if not impossible in practice.

Closest I've come is using z-score summed indexes, but generally have always found them hard to explain, and am worried about the trade off of, say, 100 observations is worth one reportable injury.