Autononomation is a critical element to lean manufacturing. In this guide, we’ll describe what it is and how it can benefit your business.
But first, you probably know by now, here’s the quick answer:
What is Autonomation? Autononomation is an important lean manufacturing tool to help increase right first time quality. It’s often called Judoka and involves identifying defects at source, stopping the line and fixing them so they don’t return. This whole process ensures that quality is built into processes and error proofing is part of your organisational culture.
There are a number of components that go into autonomoation. We’ll delve deeper to see if your business stacks up.
Table of Contents
Here are some fast links to the rest of the article…
We’ll look at where it came from, why it started and how it works on production lines.
We’ll explore the 4 phases of autonomation, which drives quicker problem solving and improved quality.
Can we use its concept across other processes? We’ll explore more.
How your business can use it to drive quality improvements and cultural change.
A Little Background to Autononomation
Historically, autonomation was created so that machines could automatically:
- Identify errors when they happen at source
- Automatically stop the line when that defect has been detected
- Allow operators to fix the problem so it doesn’t happen again, before carrying on
Autonomation is so important, it has its very own pillar in the “House of Lean,” where the emphasis is to build quality at source.
Origins From Automated Machines
Autonomaion originated from Toyota, whereby they ensured that machines were automated to the point where they could instantly detect a defect coming off the line.
An alarm would display and the line would stop. This then allowed teams to fix problems, before starting the machine again and carrying on with production.
The term ‘autonomation’ grew from this foundation.
In autonomation, machines are automated and conduct the repetitive tasks that operators would have otherwise had to do. This leaves the operators to monitor the process to ensure what’s coming off the line fits desired quality and that everything is running well.
It’s often said that autonomation is a mix of automation and human intelligence.
It’s the human intelligence that’s needed to identify opportunities for improvement and eliminate root causes when defects have been identified and the machine has stopped the line.
For more on its origins, lean.org has a good detailed description.
The Benefits of Autonomation
Here are some benefits and contrast of autonomation in relation to simple machine automation:
|People||Manual processes are easier but human supervision is still needed||Operators can multi-task. Productivity improves because there’s less reliance on humans watching the machines|
|Machines||Machines keep running until the stop button is pushed||Machine detection of defects is automated|
|Quality||Defects can be produced in large quantities because of no detection capability||Machine auto-stops, therefore crashes and issues are prevented|
|Error and Diagnosis||Errors are discovered later down the process. Root cause problem solving therefore takes longer||Errors are discovered quicker (at source) and root causes identified faster|
(This table was cited from Slideshare.net)
The Phases of Autonomation
The concept is pretty simple, as we’ve already seen.
Use automation to ensure operators can supervise machines. Automate the error detection process, so the machines spot issues and stop the lines.
Then create clear policies and procedures to enable operators to react and overcome these defects.
Train the operators in problem solving skills, and you’ve got a system that can spot issues, overcome them and ensure root causes are identified and eliminated, so they don’t return.
This is what autonomation is all about.
Here are the 4 main elements to it:
Whilst autonomation is more straight forward and easier to implement using machines, its concept can and should be used across every process.
If you think about it, how better off would your business be if all processes were designed to spot errors the instant they were created, and then systematically fix them?
From Sales Orders, to Data Entry, to Planning and of course, Manufacturing…
In time, you’d have a well oiled machine that churns high quality and highly repeatable processes, with little, or less issues than before.
It’s powerful stuff.
But yet, it isn’t really addressed by most businesses.
Let me ask you a few questions:
- Do you have a stop the line policy?
- Do employees know what to do, how and when they find defects in their processes?
- Do they implement root cause problem solving?
- Is everyone trained in error proofing techniques?
If you’ve answered no to at least one of these, then there’s a case for implementing autonomation principles in your business.
Autonomation – The Forgotten Pillar
Companies talk about continuous improvement, but do they build their production on these 4 elements of autonomation?
Do they use it in administrative processes too?
The reality is often, no.
It is pretty hard to do. when you’re not staring at an automated machine, it’s easy to miss errors. The result being; they’re often passed on to the next step in the chain of interconnecting processes.
Sometimes, this defect doesn’t get picked up and it’s passed onto the next process.
It may even travel all the way to the customer.
Other times, problems may get picked up in a receiving process, but will be corrected, without mentioning it to the supplying process where the defect came from.
The following are examples of this:
- De-burring something that should have been done at the previous step
- Touching up blemishes before you begin your work
- Entering data fields because sales haven’t entered the full amount
- Correcting paperwork just so you can start what needs to be processed
And so on.
In fact, I call these turn-backs.
A turn-back is either information or product that is passed on incomplete or not to the exact requirement of the customer in the next process step. This means they have to turn back to correct the problem before getting on with their job.
A turn-back then, consists of anything not right first time in any given process:
A missing data field; not enough parts; incomplete information… these are all turn-backs. They force you to turn back and correct the process, rather than getting on with what you planned to do.
All these elements combine to give you a bit of a corporate headache – there are so many problems from so many angles, it’s often hard to know what to do.
Let’s look at the 4 steps to autonomation in a little more detail and see how we can extrapolate this concept to more than just machines, which it was originally intended for, so we can build quality into processes.
1. Identify the Defect
This is pretty self explanatory. the first thing that needs to be implemented is the ability to see when something has been processed incorrectly.
Like our machine automatically stops when it detects a defect, we want to design our process to make it easy to see when a defect has been made.
Ideally, they should be error-proofed, so they can’t make a mistake.
Examples of this could be:
- Jigs that hold a part only the correct way
- Torque wrenches used to ensure fasteners are tightened exactly to specification
- wire locking nuts, so they can’t come loose in the field
- Adding compulsory fields before an electronic form can be sent on
- Ensuring a database lookup verifies the data entered is correct
You may have more examples.
The truth is, not everything can be error-proofed. But you can go a long way with implementing this mindset and challenging existing norms.
And then there’s the ability to see errors when they happen. Again, your current processes must be challenged.
As before, it’s not easy to do, but it reaps rewards.
Here are some ideas:
- Make in smaller batches (preferably in batches of ones) – The less work in progress there is, the easier it is to see errors
- Ensure the work area is visual and the principles of 5S are being followed. A visual workplace makes it easier to see anomalies
- Use check-lists at critical to quality points in a process, to ensure vital steps are followed and critical to quality points are cross-referenced
- Make changes to suit what the next customer wants. For instance, instead of giving a project manager a 55 page report, perhaps they only need a one-page summary sheet. Make it clear to see what they need to see and where the errors are (if any)
- Deploy targets – For instance, a manufacturing process may have upper and lower control limits to track ongoing quality. Is it within limits? An admin process may have something tracking time of response to customer or specific fields that must be completed in a database.
2. Stop the Line (or Process)
Again, when machines are automated and designed to identify a defect, it’s easy to stop the line.
Taking the concept further to other processes, it is relatively straight forward, too.
What’s needed is the ability to know what to do when the line stops. This means that there may be a number of layers to this. For instance, for critical processes (admin or production), you may have a simple policy to follow, like the example below:
- Level 1: Process operators stop the line and fix the problem within 10 minutes
- Level 2: If the problem can’t be fixed within 10 minutes, it’s escalated to the next management level to help fix
- Level 3: If this can’t be fixed within 30 minutes, it’s escalated to senior managers.
The above is just an illustrative example. But this approach ensures that there’s a timely target to keeping your processes going and getting things fixed. It also allows the business to know who’s responsible and when to get involved.
3. Fix the Immediate Problem
Whilst the machine or process is down – and the line stop policy is working – our priority is then to contain the problem.
This means ensuring that the problem is now being worked on and there’s a workaround which protects the customer.
Here are some examples:
- Keep the process working and inspect 100% of all items until the problem’s root cause has been fixed
- Outsource parts to another supplier, while the process is down and being resolved
- Inspect the process more frequently during this phase
And so on.
The point is to control the problem while you’re working on the solution. It’s like a band-aid. Keep limping along, while you get time to fix it for good.
4. Investigate and Correct the Root Cause to the Problem
This normally takes a little longer to fix, hence the need to often apply a containment action to bide you some time.
At this point, though, it’s about conducting root cause analysis to drive deeper investigation to highlight root causes.
We’re simply asking why what happened, happened.
As part of this, we use data from our target tracking, machine performance, and process performance, as well as simply observing the process as it’s running.
The idea being that it’s not fixed until we can categorically know that we’ve fixed it by eliminating root causes.
Simple tools here would be:
- Fishbone diagrams
- Cause and effect matrix
- Why questioning (often called 5 why)
- Even an FMEA
The important thing here is to ensure that everyone is trained in problem solving techniques and also given the time (cue the stop-the-line policy again), and know-how to identify a problem and fix it.
If we didn’t complete this last stage, we would just be forever stopping the machines or our processes without ever addressing the root causes.
“We Just Don’t Have the Time”
I don’t buy this excuse. You’re too busy fire-fighting to eliminate the root causes to problems.
So you keep fire-fighting and just containing actions. It’s a false fallacy.
It reminds me of a story in Steven Covey’s Book, “The 7 Habits of Highly Effective People.”
A guy’s walking through the woods and hears a woodcutter swearing and cursing. He walks up to him and asks what the problem is.
The Woodcutter replies, “The blades on my saw are blunt, and it’s taking me forever to cut down these trees.”
The guy replies, “Why don’t you sharpen the saw?”
To which he replies, “I have no time…”
If you have no time, focus on the top 2 or 3 problems and ensure you fix them for good. Take them to phase 4 of our autonomation roadmap.
Then move onto the next high priority issues in logical sequence, until you can see the wood from the trees, so to speak.
But to stop at containing problems means you’ll be fire-fighting forever and always cutting trees with a blunt saw!