Poor workplace democracy can hinder innovation and health and safety, union reps say

Published date
A hulking factory that towers over a town in South Wales might not be what first comes to mind when you think of a technological transformation. But new technology and data driven processes that affect both the product, and the workforce are well underway. Though more familiar problems persist.

In the fourth in a series of focus groups, Wales TUC - supported by Dr Juan Grigera from Kings College London and Adam Cantwell-Corn from Connected by Data - are investigating how workers in a range of sectors are responding to digitalisation and AI at work.

This write up of a session with manufacturing union learning reps is intended to capture the key comments of this meeting, which will later contribute to a full report. Names have been changed protect anonymity of the workers and to enable candidness.

An automation, health and safety and employment paradox? 

The reps acknowledged that new processes delivered significant benefits in terms of modernising dangerous processes in heavy industry, with serious accidents not uncommon.

For example, automation processes are helping to put distance between workers and the scorching processes. As E put it, the new machine “takes the operator away from the firing line” of, adding that “before its introduction, boys would load it up like a chariot. Archaic.” 

Data driven systems and AI are slowly working their way into the plant’s workings but are confronted by the challenges of ageing machinery and lack of investment, with some plant originating from the middle of the last century. Automation is also bringing with it concerns about deskilling and job losses, with jobs reportedly lost by the introduction of robotics and advanced data analytics.

However, the machines aren’t always right. As B said “if the system encounters a problem, it then asks staff to take over from there. Any time it had something it didn’t know it would say ‘you can have it back now’”, underscoring a significant reliance on human involvement.

But it's not just machine limitations that are impacting the roll out of new technology. It's also human error.

For example, one worker recounted the almost comical story of the installation of a new operating desk for a complicated piece of machinery. He says “They asked a member of staff to sit at the operating desk in the pulpit so they could design it around him. However, he was 6′ 10″. "Now I can't use the operating desk without standing up!”

This example was emblematic of concerns that the workforce was not being fully considered, or drawn on for knowledge and insight, in the plant’s management.

There was also frustration that the company was not investing sufficiently in new technologies.“The company tends to go for cheaper options,” one said. “There's one example that we often refer to of a ‘£1 million downgrade’.” 

“Directors don't even know what the future is for the company,” said one. “Staff can see the lack of investment,” he said whilst reflecting on the uncertainty which was affecting the plant and the careers of those who worked there.

Where the company had invested in AI in a central process, workers felt that it wasn’t working as effectively as it could: 

“Currently not all our sensors are accurate. So they are not always sending a messages effectively to the central computer,” said one member of staff.

One worker summarised the situation as follows:   

“I think the moral of the story is that our plant isn't reliable enough for artificial intelligence. Our plant doesn't do the same thing reliably.  No equipment here does the same thing from one ten-minute period to the next. You’d have to put so much input to deal with all the different scenarios, I don’t think it would work.”

These comments show that workers understand the potential benefits of automation and AI but in practice see that there are difficulties in adding AI to existing machines, especially when it is running a complex process on an ageing plant.

In short, these workers seem to be between a rock and a hard place: automation and AI come with the risks of redundancies and higher levels of surveillance, but opposing to any form of investment puts the plant itself at risk of obsolescence. 

For example, the workers cited the roll of out of swipe cards for moving around the plant, which was accepted by most as part of safety and security but concerns we raised that it would be used for surveillance and further intensify work.  

Worker input on workplace management and new technology

This concern is not just from the firm’s workers. A common theme emerging from these Wales TUC convened sessions has been the lack of participation by the workforce in the roll out of new technologies.

Some of the participants reflected how this had changed over the decades they had been at the works. A said workforce participation “pre-1995 was good. When new machinery was installed, staff were given extra money as their jobs required more training and more skill. You’d have been taken to wherever the kit was already in use. The unions, the operators and the managers would have looked at it together. And the unions would have negotiated a pay rise on the back of needing extra skills to operate.”  

But now such co-determination seems to have weakened A said. “New technologies are always sold to us as if the management is doing us a favour. Normally a new technology comes in under a reorganisation. The union is told about the new equipment and we may be told that in time it will take a person out of work. We are pushed down the health and safety route in terms of being persuaded to adopt new machinery.” 

Or as B put it, “when new software is introduced you're told ‘this is it’”. Shop floor workers are not involved at all, while union officials might have more of a say. “There’s no engagement really, workers are left alone on the shop floor. People feel undervalued. Sometimes a new system was put in and we were told to get on with it. Until it’s pointed out to them that it’s breaking health and safety law” B said. 

Exacerbated by remote meetings with management, the worker concurred that this led to poor industrial relations. “They just sent us new policies” Barry said. “The old school HR directors aren't there anymore. And there is no consultation. The art has been lost. We end up arguing with them then and there’s friction.”

A growing distance between the workforce and management tracked across other areas of the plant too. According to the workers, it was weakening health and safety reporting, as well as exacerbating inefficiencies.

For example, a previously simple system for reporting and escalating accidental ‘near misses’ has been replaced. “The new system is semi-automated on a computer. It prompts you and there are drop down boxes.  It determines the level of investigation according to how serious the incident was. But it is too complex and it puts people off. We have criticised the system over and over again at the health safety committee.” 

As someone else put it, the collection of data seemed to be the top priority, rather than reporting health and safety issues. “Data is being collected but the new system isn't working for shop floor workers. It’s not making things safer, but it is making their life easier.” There are now serious concerns about under reporting of health and safety, something that could have been avoided if the workforce had been brought into designing the system the workers said. 

Similarly, a suggestion scheme to improve work processes has been semi-automated and made so cumbersome that workers feel disincentivized to propose how the plant can work better. Monetary incentives that used to be significant if proposals were adopted have also been reduced to a minimum. 

The workers saw this as a disregard for the value, trust, and knowledge they could bring.  

This has also seeped into workforce monitoring and HR practices. “The risk is that we are losing the human elements in HR” through automation and remoteness, "there is no benefit of the doubt, no grey areas". Adrian said “I know we are told that computers can’t get it wrong, but they can get it wrong.  They are just doing what they are told. There’s no human element in there.” 

Someone joked “I can picture the scenario now. A union robot representing a member against the HR robot!”. Without missing a beat, someone else quipped “It wouldn’t look that much different from now!”.