Skip to content

Good work design and human factors – employee engagement and safe design for a parcels handling system

Musculoskeletal Disorders Symposium 2017

Stephen Hehir

Presented by: Stephen Hehir (Australia Post)

Run time: 36:03

Download a copy of this podcast (MP3, 21 MB)

Presentation 8: Good work design and human factors – employee engagement and safe design for a parcels handling system

Presented by: Stephen Hehir (Australia Post)

[Start of transcript]

Nita Maynard:

Alright. We might kick off. Thanks everybody for hanging in there and coming along to Stephen's presentation. I think it's going to be excellent.

My name is Nita Maynard. I'm the Director of the Work Health Design Branch with Workplace Health and Safety Queensland. So you are in the good work design stream, which is fantastic and very exciting.

So our next presentation is Stephen Hehir. Stephen has over 20 years' experience working as a specialist HR consultant. He has experience in navigating the work health and safety industrial and regulatory maze, enabling the right people with the right behaviour and the right system to get the right safety outcome.

As part of a $2 million future network renewal, Australia Post invested $595 million to build two world class facilities for their parcel handling. With safety as a key cultural pillar in the corporation, safety was comprehensively considered from the initial scope of the future parcel network program, through design and testing, to the implementation of a new sorting automation and system of work.

This presentation describes the process followed to ensure a safe design, through to realisation in an environment requiring speed of action, accountability, collaboration and customer focus. Consultation with the workforce and customers has resulted in significant innovation. The case study focuses on employee consultation and the ergonomic aspects of safe design.

Please join with me to welcome Stephen to the podium.


Stephen Hehir:

Can everybody hear me okay?

Can you hear me okay?

Not really? You can now? Okay.

Okay. Well I've done today more as a case study because it was really about giving people an opportunity to see what we did – how do you operationalise a lot of the theory that we've discussed today.

So what I've included in this is the actual packs that we had in our consultation process. Now there's a lot of slides, and I'm just going to state that upfront. I wouldn't normally use this many slides, but in order to show you what we actually did and just flick through, give you a bit of a sense of that, I thought I'd include all of those slides in the deck today. So a lot of it I'll kind of speed through. I'll just touch on some of the key points. But it will give you a sense of what we actually went through in that consultation process.

Safe design, it's nothing new. It's about thinking through what might go wrong upfront, having a sense who are the people who will be undertaking the work or who are the people who might use your end product or who are the people who are going to be driving across your bridge. So it's thinking that through.

And it also takes into consideration how you're going to build it. It's one thing to have the product, the bridge, the machine built at the end, but the people that have to put it together in the first place, you need to think through their safety as well. That's a major issue in a lot of building and construction, and when you see some of the size of the plant that we've built, I think you'll get a bit of a feel for it there as well.

So you've got to consult widely. It is cheaper – you'll all find that it's cheaper to do this design piece upfront, get the safety built in upfront, than to try and retrofit it at some later stage. And I guess a lot of you have come across that. We used CAD, computer added design, fairly extensively to walk through what we'd built, put people in there and actually have a look and see how do they fit so to speak. We looked at the incident data. We got a lot of information about the products that we were going to be sorting, what I'll refer to as a dispersion. That's the weights of parcels that we're handling – this is a parcel sorting machine – and where is the majority of the weight in terms of that profile of weight distribution. It's very skewed to the lower weight products.

We also have a focus at the moment on what we call serious injuries and fatalities, or SIFs, and you might remember in one of the presentations this morning on the metrics, we're used to seeing LTIs at the top of the pyramid and the first aid cases at the bottom, and we're sort of tipping that kind of on its head and having what are those things that are going to kill you? What are those things that are going to cause life altering injuries? We put those at the top of the focus, because they're the things that if you can get that bit right you're going to make a bigger difference.

And based a lot of decisions on evidence and also the maintenance. We put a fair bit of effort into thinking through what are the common maintenance tasks and how are the technicians actually going to do that routine kind of maintenance. That can often be a big issue for people.

I put a little side note there just to have a look at the issue of older workers. If you can design things – and I'll say that upfront, if you can design things right for the older worker, you're going to protect everybody along the line. As I said, nothing new about safe design. If you want to write anything down about my presentation, write down Safe Work Australia's website and go to their safe design section, because most of what I'm talking about is in there. This is nothing new. I'm just putting it into action.

So the sort of things they're talking about you can read on their website. I'll just bounce through those to save a bit of time. The principles are very interesting to think through, sometimes a little harder to achieve, like the lifecycle piece. Sharing the knowledge I think is important in there. But anyway, if you have a look at Safe Work Australia's website you'll get a good sense of those.

At the time we were doing the design work, this is a bit of a look at the manual handling injuries, the claims by age if you like. So as you can see there, the younger age groups seem to have more of the injuries. If I had a look at that by age in terms of days lost, which is a severity measure, what you can see is that the older workers have the more severe injuries. So it's like the younger people have more incidents, but the older people, box clever, don't have as many incidents but when they do they're doozies. So if you protect the older workers you're going to inherently protect everybody else at the same time, and that's the way we try and look at that. You can see those two charts there together.

This is the ground floor view of the Sydney plant. It's bigger than the MCG, and this is just ground floor level. They're three levels high, and we'll have a bit of a quick look at a picture of that. I'll get a mouse down here. So there's a lot of forklift activity there, and we'll show you a little bit of that, and we've used some other equipment to move – ULDs, unit load devices, in the north south direction. So that's safe design actions.

Getting the right people involved early in the piece. So the design team when they had appointed a project director, the first person he pulled on was a safety person, which was good. Including safe design requirements in the engineering specifications if you like. Now that's a 52-page document that's got a lot of requirements in there. Our workforce is quite familiar with that through years of industrial participation and consultation. They're very familiar with the kinds of specifications that we put up around safety.

Consultation and engagement, it's kind of key. If you're going to make it work, you've got to bring these guys on the journey with you. You've got to engage them. You've got to get the buy-in from them about what you're trying to do. You need to understand from their perspective what's an issue and how do we deal with that. Bring them on the journey and you'll have some great success.

One of the successes that I had, at this Sydney site actually, they wanted to make some shift changes after the project, and they'd learnt this technique using NIOSH. And they'd gotten together and kind of worked out a bit of an argument of how they could change their shift arrangements. The only reason I got called in was they, being the HSRs and the staff, were trying to remember some of the intricate details of the NIOSH tool that they'd used. So that's where we got to.

I'll talk more about that later on. Collaboration with the vendors, getting them on side and understanding what we're trying to achieve. Don't just leave it to them to come up with a design. You've really got to drive what you're after. You've got to put your DNA into that design.

Have a look at what's happening in industry best practice. We went around to a lot of other mail organisations and had a look at their equipment, and we didn't just take our own word for this, we brought in some third party experts to check off the ergonomics and the plant safety side of things.

I'll see if I can make this work. This will give you a bit of a sense of the scale and size of the space.

That's from the second floor. At what is called a cross belt sorter. This is a cross belt sorter. These little belts here are like a little conveyor belt that can run left or right and push a parcel over into a sorting area. And you'll see the conveyor belt tracking along. Just watch, I think it's that larger parcel there.

No. It's the one on the other side. See how that gets pushed off.

So that's what a cross belt sorter is, and that's how that works. In terms of the load shifting equipment, these places are massive and there is a lot of forklift activity. This is some modelling we did to show you the forklift movements. Now would you want to put any pedestrians within a bull's roar of all that fork truck activity? You know, what are the things you can do to absolutely keep people out of that area? In the 20 years I've been at Australia Post we've had five fatalities. Three were motorbikes and two were forklift related. And on another occasion when a gentleman in Queensland lost his leg from being struck by a forklift truck, I had to front up in the Federal Court over that one. So I mean fork trucks to me is the big thing to deal with. I don't know what in Queensland – in Victoria I know it's the case that most industrial fatalities are fork truck related. I don't know whether that's the case up here or not. Outside motor vehicle crash that is.

So we took on some aspirational goals upfront. Reducing manual handling by 50 per cent was one of those. We adopted a direct load to ULD way of managing that. So manually just lifting parcels the whole time, we took 50 per cent of those parcels out of that lifting equation and found something more useful and productive that involved decision making and other things for the staff to do. So we had the same number of staff, just that we reorganised the way that the facility worked.

Sydney sorts 35,000 parcels per hour in that order. So these are massive machines, massive processes.

Eliminating the manual handling. So you get a bit of a sense there if we put our containers, our unit load devices – each one of these cages that you can see down here, think of that as a van driver, the guy in the little van that comes down your street delivering the parcels. We're sorting to that level. That's what we're trying to achieve with this. And we had a look around the world to see what other postal organisations did. Finland – if you're into following quality education, you'll love the Fins. I really enjoyed visiting Finland. They were the only other organisation that did direct load, and they were dropping up to three kilogram parcels through about 3.6 metres and sorting. We asked them, 'Don't you have any problems with breakages and things?' and they said, 'We don't have any customer complaints'. They had a couple of dints in the outside packaging, but they weren't having any problems.

We were conservative. We said let's try one and a half kilograms and let's maybe drop only half that distance. Quite frankly we could have gone much higher. We also said 50 per cent upfront. The experience now I can tell you we could easily have gone 70/75 per cent of our parcels being totally automated in the handling. That's the way that that went. So that's down the bottom, some robotics that takes those ULDs, the unit load devices, those cages that you can see with the parcels in them. There's a robot that takes them away behind when they're full. So the people are doing the decision making. Has there been some error made here? Is the cage full? Do I need to stop and fix something up here? So it's using people for what they're good at rather than using them as a machine. You can see the sort of railway track here. It looks like a twin fork truck system that goes up and down a track underneath the sorting system to address those. So that's how we got rid of half of the sorting.

And that's just a picture to give you a sense of the scale of it. I think it's about 380-odd van sorting rounds if you like, and there's 380 cages in that centre.

We wanted to eliminate lifting at what's called singulation. So parcels delivered to this area here, in the past what we would have to do is actually pick the parcel up usually and turn it over so that the barcode on the label was visible to the machine. So there was a lot of picking up and rotating and repetitive movement. So we had a look at what people overseas were doing and we found in Sweden that they were putting a camera underneath and had a gap in the conveyor belt, and they were reading underneath. So we were able to then get six sided scanning of the parcels to read the barcodes. So that got rid of that lifting.

This just becomes a movement activity where they're just pushing one parcel through at a time, and they swap sides every half an hour. And they also have a job rotation to other different types of jobs. So this is a non-lifting task. I'll come back to that later.

Application of the NIOSH manual handling principles that we heard Robin talking about before. Just a caveat, it's quite specific to lifting. There's other tools that you can use that look at pushing and pulling and grasping and other things. There's 3D, the University of Michigan program. And also it is one of the tools that's recommended within the model guidelines for hazardous manual tasks. That's what it looks like, and you can see where the NIOSH tool is included on there.

Now there's a fair bit in this. It's not that straightforward to use. It's a 164-page manual. It requires a fair bit of commitment to learn to use if you want. And it's catering for 99 per cent of the male worker population and 75 per cent of the female worker population. I'm going to suggest to you that if you want some help, have a look on the Ergonomics Society's web page. You can see where there's 'Find a certified professional ergonomist'. Most of the ergonomists being trained in Australia at the moment are fairly au fait with using that tool if you don't want to sit and read 164 pages.

It's an equation that looks at the kinds of things Robin was talking about earlier today, the vertical lift height, the horizontal distance from the body, how much rotation of the body is involved, and probably the most sensitive factor, the thing that's most influential, is the frequency of the lift and the duration of the lifting task, and you can vary those a little bit. Now there is an app you can get. Just put NIOSH. That blue thing will come up. It's about $1.99. You can punch in the various items on there and it will give you a solution. For instance with the posture I'd put in here looking at a 16 kilogram parcel, it told me that it was safe to lift up to 17 kilograms in that posture. I'm trying to keep this lifting index below one. So that's the way that sort of works. And if you want some expertise in that, then go and have a talk to one of our professional ergonomists.

I'm giving a plug there, because I'm Chair of the Professional Affairs Board of the Ergonomics Society.

So the most sensitive part of the equation is the work duration and the frequency of lift. If you're lifting something every five minutes, go and knock yourself out. You're probably not going to have too many problems, unless it's a ridiculous posture. But by the time you get up to six parcels per minute and you're rotating every two hours, we've just taken half the safe weight off you. So the work rate is a very influential part of NIOSH. If you've got a job there that requires you to do some repetitive lifting and you're up around 13/14 lifts a minute, the safe limit that it will give you is going to be fairly low.

So let's apply this. I'm here in the back of a semi-trailer. This is what we call loose load and loose unload. And if you can imagine a conveyor belt is extended into the back of the truck. You can see me standing there to demonstrate this on an elephant footstool like you'd find in a library. The only lights in there are attached to the end of the conveyor. It's actually quite dark up in the top. Things fall on people, and there's a lot of overhead work. Hands up those who don't think that's such a good lifting job to have? Yeah. Okay. So let's throw that in file 13.

And the only other place I saw this – and they didn't have the vacuum lifters but they did have the – they call them man rider or person rider platform. So we put a platform on there that the person actually stands on. Like something out of Thunderbirds from the 1960s isn't it? And you can see what influenced me. But I'm okay with that, because it also influenced NASA that program. And then we added the vacuum lifted gantry, which we'd seen being used in a very different application. So we married the two together and we came up with a solution where you can lift in this case – it's rated to 35 kilos. That's a 20 kilogram box that you can lift with that vacuum lifter. So there's no lifting involved.

So this platform extends into the back. You can raise it up and down. You don't have to stand on an elephant footstool. We don't filter within 300 mils of the roof, so all the lifting is with your upper arms below the horizontal, and we control the lifting rate. The staff love it, and they love the opposite direction where it's like a game of Tetris, having to put the parcels in there. They really enjoy that particular task. So that was our kind of major innovation.

Vaculex who make the vacuum lifters sent a team of six engineers out here to have a look at what we did, because no one else had done that anywhere. Because they could see that this was going to take off, and it is. DHL in Belgium came out here to have a look at it as well, and now they're implementing it. So it's the kind of innovation that we will see in the future.

We have this recommended weight limit on this manual induction process here. It's around the 10 kilogram mark. So if a parcel is marked at greater than 10 kilograms or it's got a 'Heavy' sticker on it, they use the vacuum lifter for that.

I'll race through this. It's a little trolley that you can pull on that's height adjustable. Maybe I can quickly show you.

So if you've got a heavy product and you can't lift it, put it on the trolley, lower it to the right height.

And then just push it in to the unit load device. So that's the trolley there. Four pumps and it lifts back up to the top height. It probably gets used one out of every 100 parcels that's exceeding the recommended weight limit. We were really pushed for space there. We just couldn't put any other kind of mechanical handling aid in, and the staff were okay with that trolley concept. They're now looking for one that raises – they don't like the electrics. They're too slow. But they are looking for one that will raise and lower a much greater height range than what we had.

So our injury rates. In the similar facility it was about 34 manual handling related injuries in the 12 months prior to the implementation of the new system. We had four afterwards, and they were fairly minor. And there was one lost time injury which was a broken finger where a parcel got dropped on someone's finger. We've had no industrial disputation, which for us is pretty good. We have quite a heavily unionised workforce. Positive feedback from the staff.

Probably one of the main issues I get anxious about – Kirsten Way's talk before about diversity in a workplace and so forth – we have a social workplace. People historically came to Australia Post, they'd sit in teams and they'd have conversations – it was a very social workplace – while they sat and sorted envelopes. So we've tried to maintain that, and in order to achieve that we need sound levels down around the 75dB do you don't have to scream at somebody to have a conversation. Now this machine that we ended up with, even though we had a specification to 75 and they promised us that they'd give it to us at 75, quite frankly it's much louder. It's about 81. I'm not happy about that, and it has interfered. So we've ended up putting a lot of acoustic damping foam around to try and reduce the noise levels at the workstations, not because it's causing a hearing loss problem, but it's interfering with our social workplace. So that's important. So that social element of work we've tried to maintain.

A quick look at those aspirational safe design pieces. Segregating pedestrians and forklift trucks. You can see here, that's where the fork trucks load something up. This is where a pedestrian pathway was in the old facility. That's just not a good look. And you can see this, a forklift aisle way down here, but the pedestrian line which you can't quite see anymore, that's where the bosses used to walk across the floor. And you've got fork trucks cross this way and pedestrians go down there. That is just not a good mix. If you go to Safe Work Australia's guidance on segregation of pedestrians and fork trucks in the Traffic Management Code of Practice I think it is, you'll see this use of overhead platforms and putting in the amoral kind of stuff. And that's what we've ended up doing. We've cut down nearly half of the fork truck traffic by putting in these smart conveyors if you like.

This is in the Vechel Test Loop over in the Netherlands.

Sorry. Not working. Move on. These conveyors slowly move through the facility in the north south direction to take these unit load devices or ULDs around. So that way you don't have to have the fork trucks in there doing it, you've got another means of doing it. And it's quieter. These things don't make any noise. Fork trucks are noisy, beeping reverses, flashing lights, people turning their head around, you know, sore necks after a while, that sort of thing.

I wonder if this one will show. This is a de-stacker. You can reduce your number of fork truck movements by lifting two ULDs at a time, and then you have a device that unstacks the two of those. So it's not a fork truck making – you halve the fork truck movements by mechanising the number they have to make.

We wanted to prevent falls. Falls from height was a SIF risk. It's the serious injury. It's life changing stuff, or in most cases it's a fatality. So we ended up putting in 8,000 square metres of platform in Sydney and 5,000 square metres of platform in Melbourne. We added the little crossing bridges to that platform for pedestrian movement purposes, and so there's a heck of a lot of platform up there. And that's the way we got around our fall from heights risk. So instead of the techs coming along with boom lifts and ladders and all this other stuff to do maintenance work, they've got platforms and they love it. And talking to them about, 'Guys, where do you want it? What are the things that…' – and they said, 'I don't know. We really need it everywhere'. Well the budget was not extending, so we kind of agreed on what are the 80 per cent of places that we can put it in that you're going to be using most of the time, and they were happy with that and we came to an agreement about where to put the platform.

So employee consultation. These are the decks. We talked to them about some of the safety issues, fitting people – how am I going for time?

Nita Maynard:

A couple more minutes.

Stephen Hehir:

Okay. We talked to them about people fit, anthropometry, we talked about biomechanics, energy expenditure, where to put the controls and displays, involved them in decision making around that. Involved them in decision making around work rotation and which parts to rotate to and from, and we discussed in detail the seven workstations.

So the anthropometry, we tried to get them to understand those people differences and why workstations had to be built in the way that they were. Have we biased it somewhere, got more shorter statured people in one part of the business than another? Thinking that through.

The NIOSH stuff, we took them fully on the journey with that. We looked at energy expenditure and had those jobs assessed, or we did that with them. We talked about singulation and they came up with a number of ideas around that workstation, the kinds of design changes that we made which we incorporated – little roller balls and things on the side.

This manual load infeed area, we ended up coming up – apart from the vacuum lifter, they wanted these elevated so you didn't have to bend down so low. So we used a BT lifter, and we bought a heap of them and shoved them in, and staff loved that so that's what we ran with.

You saw the direct load workstation and we talked through the detail about that and what things might we change and what are the kinds of issues you might have there.

So they were the sorts of things. All of the control systems. They thought rather than me having to – I get a parcel, take it to ULD10. Where's that? So they wanted a flashing light to show them where that was. So we were able to make some software changes where the barcode scan actually lit up a light and they knew where to take it, they didn't have to be cognitively drained and make errors a lot.

Some of the different workstation arrangements. We talked about this. Where do people stand, what might you trip over and what might you bump into? We got their feedback, and they were interested in what have I got to see? What visibility do I need? Can I see the TV screen that tells me how we're going and all this sort of stuff? So there was a lot of discussion around that. And that's the loose load conveyor with the platform on it. A lot of conversation around that. Ninety per cent of the parcels under 7.1 kilograms, so it's about one in 10 that they would end up having to use that vacuum lifter for in that process.

There's the dispersion there. Interestingly, parcels coming through our facility median weight, around the two kilogram mark. With more and more lipstick and little things being purchased overseas, that's actually dropping down now to more like 1.8 kilograms. A 32 kilogram item is a one in four million event for us. So that's people's perceptions about the weight of things.

So that's the sort of stuff that we work through. Just to maybe highlight one or two others. The man machine interface as the Dutch call it – they were keen to sort of point out to us that if you can reduce the work rates a little you can actually lift the amount of weight that people can lift.

So they're the sorts of things that we talked through. I'll race through these and see if I can find the – there were two categories of tasks, lifting tasks and movement tasks, and we would rotate every two hours. And that looked like that. So they'd sort of rotate from one type of task to another type of task. And in the lunch time, sometimes we actually made two – instead of having a quarter of an hour tea break in the morning, half an hour for lunch and a quarter of an hour in the afternoon, they changed their working arrangements to have a 36-minute extended morning tea and a 36-minute extended afternoon tea, no lunch breach, and then they would spend two hours in each of the tasks in the in between time. And that meant they could go from a lifting task back to a lifting task, and they came up with that themselves.

I might leave it there.


Nita Maynard:

Thanks Stephen. That was fascinating. I had a quick question. Did you utilise mock-ups at all so that people could actually try things out?

Stephen Hehir:

Yes, we did.

Nita Maynard:

And what were your learnings on that?

Stephen Hehir:

So we did make some mock-ups in the Netherlands before Vanderlande went to production on the workstations, and we took HSRs from each of the facilities and some union reps along as well to the Netherlands, and we worked through that with them. We got those changes made, some of which were they didn't like the idea of having to reach to get certain things, so they put these spreader dividers on the singulation workstation. They edge was uncomfortable they felt on the indirect load units, so we changed the construction of the – we put timber sides on them basically, and we ended up putting timber sides on a lot of the workstations and put roller balls where they would have to drag some things. So they were just little things, but they got quite interested in – we just let them go basically, and the Dutch engineers just sort of took notes and did the little things that they wanted.

Nita Maynard:

I can imagine that would be a critical part of your consultation and engagement. And it's very hard to visualise things. It's often very good to have something tangible that you can try out. Yeah. Okay. That was my question. I was selfish there. Is there anyone in the audience who has a question for Stephen? Yep Vicki?

Audience member:

Is there much of an error rate from the scanning?

Stephen Hehir:

The challenge I guess that we've had is we've integrated Star Trak and Australia Post, and they've had different systems. And we've purchased the Vanderlande equipment and the Erwin Sick instrumentation, and with those changes you get some errors. So there is some error, and there's a whole army of people that are working to try and address that kind of error rate that can occur in a major project like this.

Nita Maynard:

Next over here?

Audience member:

Great presentation. I was hoping it would be available. We do a very similar task set at the airport with the conveyor belts and moving bags and things into load devices. With the conveyor systems you looked at, did you find any others of a similar nature that would throw an item from one side to the other?

Stephen Hehir:

Sorry, to throw the bag?

Audience member:

Move it, sorry.

Stephen Hehir:

Yeah. In our airport systems there's these diverters. They're like an arm that comes out to the conveyor, and it has a vertical conveyor on it, like the pinball things. It's not as rough as a pinball machine, but it does the same kind of thing and diverts the bag off a conveyor down into a chute. But those cross belt sorters are much more efficient and much gentler on the product.

Nita Maynard:

A question down here?

Audience member:

Have you seen any knock-on benefits for your local depots, your local offices, or has the environment there been mostly unchanged?

Stephen Hehir:

So what they're receiving now – instead of unit load devices that are full of parcels just for their facility, they're now receiving – which they had to then do a sort, so more manual handling. So we've now given each person their own container full of parcels for their van so to speak. So they don't have to do a preliminary sort, they just do the one sort into their van. They sequence it into the order that they're going to deliver it.

And one of the things we're sort of working through now is do we start to put shelving in, do we have to put other containerisation systems in the back of the vans and the trucks to make it better from a manual handling point of view. If anything that's probably a weakness in our system at the moment which we're working on.

Nita Maynard:

Any other questions?

Okay. Well please join with me and thank Stephen again. Fantastic.


Okay. We're now going to break for afternoon tea out in the foyer. Again please visit all the trade displays and get to know everyone. We will reconvene back in our main room at around about 3:30 I think it is. Thank you.

[End of transcript]