The Lake and the Rocks

One of the metaphors that they use in Toyota is called the "Lake and the Rocks". Is goes like this.

The lake is the work in progress (WIP). That could be inventory in manufacturing, it could be partially processed mortgages or patients waiting on trolleys in corridors waiting to be treated or admitted. Anything that represents a partially completed job.

The rocks are the problems that slow the work down or cause errors leading to unsatisfactory work.

The main idea of Lean is to make value flow to the customer as fast as possible. One thing you have to do to make value flow faster, is to remove WIP from the work stream. If you want cars to go faster on a motorway it helps to have fewer cars.

So the lake is the WIP and the rocks are the problems. The metaphor says that as you reduce the the water the rocks will be exposed. So as you make work flow by getting the excess WIP out of the system you will come across problems that will hamper further improvements.

So work on the problems and you can continue the improvement.

The trouble is there are always more rocks. That is why Lean is something that an organisation "is". Not something it does and then stops.

Eliminate Waste? Prevent Waste?

In a conversation with a colleague an intriguing thought came up. Is Lean about eliminating waste or about creating a system that prevents waste in the first place? I think that is the difference between continuous improvement (kaizen) and step-change improvement (kaikaku). Kaizen tries to reduce the waste in the current system. Kaikaku creates a new system with no waste built in.

In the end you need both. Too many trundle on with kaizen and never make the step-changes they need to. Similarly, many make big changes and then sit on their laurels for years wondering why the world is passing them by.

Standardised Work

Standardised work is bad. If you standardise everything, the work becomes boring, you can't deal with the special cases and the cost of documenting the standards is too high.

Well, yes and no.

There is no point in having standards yellowing away in a dusty folder, locked in a rusty cabinet. There is no point in sticking only to the standards like the script in the call centre that won't let you ask the simplest question until all the verification steps have been gone through. There is no point in making all the staff bored by giving them all the same work to do every day.

But that isn't really what standardised work is for.

At its best, standardised work is a way of capturing the best yet known way of working. No more, no less. I must emphasise the phrase "best yet". Just because it is standard does not mean there is not room for improvement. It is simply the best so far.

The standardisation should encompass variety. In fact some work should not be standardised. A good example is dealing with failure demand. Do not automate failure demand. But the high volume, fast moving value demand probably should be standardised.

Once the fast moving, high volume work is standardised, staff will have more time to work on either other work or perhaps further improvements, thus making work more interesting not less.

So be careful what you standardise and use it to lock in learning and not to control staff.

Let the system be itself

I was helping to run a workshop for NHS staff to assist them in preparing to meet the 18 week target set by the government. The target requires that the time between a GP referral and treatment be less than 18 weeks. The fact that targets are rubbish is not something that I am going to go into today. The fact remains that the NHS has to try to meet this target by hook or by crook.

During the latter part of the workshop an attendee raised a very interesting question.

"Is there any best practise to tell us how to monitor patients? When should we intervene to ensure that a patient does not breech the 18 weeks."

They went on further to explain that for example, if a patient has not been seen by a consultant for 6 weeks since the referral, should they bump them up the queue to ensure that patient won't breech the 18 weeks.

The problem with this thinking is that it is only thinking of the target and how to hit it. It is not thinking of the current system and its capability to hit the target. In fact I would advocate forgetting about the target in order to hit it.

In order to do what the attendee was asking about, they would have to add in a whole layer of monitoring and expediting mechanisms. This would take resource away from the patient and their journey through the system.

My answer was that, at present there will wide variation in the end-to-end time from referral to treatment and that many patients will breech. You need to understand that, but as long as the system is stable, you must not act upon it. The way to proceed is to understand the variation in end-to-end times, ensure it is stable, then work on the system to improve it. Look at the end-to-end process to improve the work on the ground until the system is predictably capable of hitting the 18 week target.

If you tamper, you will only increase the variation. Everytime you intervene to expedite a patient you displace other patients and the variation increases. You are worsening the system, not improving it.

The attendee replied that that was all very well, but that they would get "killed" if there were target breeches after the deadline.

Ho hum for the real world. Does government or the Department of Health realise what it is making the NHS staff do in order to meet their targets? If they did, I would hope they would stop setting them.