Search This Blog

Thursday, May 04, 2006

Can a benefits based testing approach be used in Safety Critical areas?

Here at TCL we have a benefits based testing approach, which we use to try and ensure that projects deliver their anticipated benefits and add strategic value to their organisations.

It’s a good technique - borne out of a very simple concept that testing, and indeed the project should be driven by a business need; that need can be expressed in terms of benefit - normally fiscal - and we should take a holistic view of the project which is led by that benefit.

This enhances and possibly even opposes the resource based logic that we would normally take through the more risk based approach.

My question today though is: Can a benefits based approach be taken to safety critical environments?

When I first started thinking about this I started to join up the idea that even in a safety critical environment there are tangible business benefits. Safety critical applications are focussed on not having defects that cause harm - and indeed that is a valuable thing to ensure is ok. Can that value be quantified though? Can it be transposed to fiscal terms? Should it be?

In the UK our culture aligns closer and closer to the US, and in so doing we see a more legal orientated approach to business. Perhaps, like Ford and their factory recall of a few years ago an evaluation could be done of the cost to make something compliant with safety critical requirements and the cost of a defect in terms of law suites? This approach doesn’t sit well with me ethically, but am I just nieve?

If we don’t take this kind of approach then I get a little lost in terms of how to quantify the business benefit of something being safety critical, how we would then link this into a business case and then in tern link to requirements. The answer is out there, but I think there is some more work for me to do before we can get this to work.

My colleague Barry is putting together a white paper on benefits based testing and if you get chance to read this it would be great to hear your thoughts on it, and indeed how benefits based testing could work in a safety critical environment.

1 comment:

Paul Darby said...

Can benefits based testing co-exist with risk based testing?

I'm enclined to believe that maybe where risk is possible, certainly in a safety critical or mission critical environment, the risk should outweigh potential benefit.

To provide an example: I've worked on a system that receives messages against fast moving aircraft, ie jet fighters etc. A benefit for the end user would be for the system to be able to decode and visualise a greater number of messages and tracks on a chart. This is great for the user and a genuine benefit to test against, but the knock on effect is that a sychronisation tool can't cope with the throughput of data and other attached servers experience a time lag.

If we went down the benefits route we may be more enclined to focus on the good plus that comes from the benefit. If we lose site of risk then we may cause more trouble than we can afford.

In summary, I believe that Benefits based testing works until a risk is identified, at which point risk based testing should take precedence.

Hope that made sense

Google PageRank

Blogpatrol Traffic Statistics