Digital Energy Journal- Making data easier to work with

Join | Existing Member Login

You are Home » News » View Article

Making data easier to work with

Thursday, February 24, 2022

How can we do more to make digital tools more useful for employees such as factory workers? Cognite put together a panel from Aarbakke, Forrester and VISMA to discuss.

While our digitalisation systems get ever more sophisticated in the processing and analytics they can do, there is a parallel challenge, making the data output easier to work with.

Cognite put together a panel to discuss how this can be done better, as part of its 'Ignite Talks' online event.

The speakers were Ole Petter Nordanger, analyst, BI consultant and specialist in data visualisation and contextualisation at VISMA, a business software company based in Oslo; Paul Miller, an analyst at Forrester Research, with a focus on digital manufacturing, AR and IOT; and Rolf Thu, Head of IT and Smart Factory manager for Aarbakke, a machining / mechanical solutions company based near Stavanger, Norway. The moderator was Petteri Vainikka, VP of product marketing at Cognite.

How good are we

The speakers were asked how close we are to a point where workers on the factory floor have access to trustworthy meaningful data, on a scale of 1 to 10.

'I think a lot of companies possess a lot of data. However, the extent they are able to utilise it, draw information and gain insights, varies a lot in my experience,' said VISMA's Mr Nordanger. 'On average I would say the companies are a 4.'

'I wouldn't go as high as 4, There's a lot of data but it's not available for workers,' said Rolf Thu of Aarbakke. 'In general, we are probably at 3. Some companies like where I work, we are a little bit higher. I would say we [Aarbakke] are 6 to 7.'

'I'm going to go even lower,' said Paul Miller of Forrester Research. 'There are some lighthouses, there are some organisations getting this right. But for most organisation a worker on the factory has no access to the insights they need, so 2.'

Getting data to people

One of the first priorities is just making data available to people, all panellists agreed.

Visma's Mr Nordanger said, 'we have to provide data to everyone in the company. If they are not able to get the data, how are they supposed to take a data driven decision.'

'Have patience, becoming a data driven company is not something you do overnight or over a month, it's a continuous process you actively work towards.'

Aarbakke's Rolf Thu said, 'for executives it is about supporting and not ordering too much. Then you need to get data to the users in a structured way. Then magic will start to happen. People want to improve their work; they want to learn.'

It is about 'getting the data to people on the factory floor who can actually do something with it,' added Forrester's Mr Miller. 'Getting them to solve problems they know they have.'

'This isn't about data for data's sake, this is about understanding pain and friction and risk and challenge in people's day to day work. Giving them the tools to fix that for themselves or giving them support to allow others to fix it.'

Data visualisations

Visma's Mr Nordanger was asked how visualisation tools can help. 'The beauty of visualisations [is that] they have this ability of compressing huge amounts of data and highlighting the essential information,' he replied. 'We humans have tremendous visualisation perception. It's easy for us to collect this information if it is shown in a proper way.'

'I believe that data visualisations serve as a tool that helps us gain insights with complex data, seeing relationships and detecting patterns. All of this helps us to take data driven decisions.'

'Manufacturing is a high paced environment. Things change, a lot of data is streaming in, existing data might be updated. Using a standalone visualization will probably not be enough. You have to connect your data sources into one model, then you can build visualisation solutions on top of them, serving as a sort of visualisation layer.'

'You have this tool which allows you to slice and dice and move into different branches of your data and start gaining insights. You can customise dashboards, so they fit the 'glossary' of the workers, so they understand what they are looking at.'

'Our customers can start looking into the supply chain, or look into the performance of assets and operations.'

'We're starting to realise the value of being able to provide intuitive visualisations.'

'There's a lot of challenges as well. Manufacturing has a lot of different data sources. Providing one seamless data model that's able to capture a lot of this information is very complex. On top of that, [a challenge is] providing a simple way to extract information to business intelligence (BI) tools.'

A challenge with data visualisations is knowing whether or not they are useful. 'We see often that companies develop solutions, but workers aren't actively using them when they take decisions. We often forget to involve them in developing the solution. At the end of the day, they are the ones that are going to use it.'

'If you're able to involve them in this [digitalisation development] process they might gradually understand and realise the value of how visualisations might contribute to their work.'

Listening to workers

'I think there's a clear need to listen to the workers on the factory floor, [find out] what they are trying to do, where's the pain and friction in that job,' said Forrester's Mr Miller.

'I've seen far too many examples of very fancy technology deployed from on high. The CIO team or the 'emerging tech' team walks onto the factory floor with this tool and says, 'here you are, this will do your job for you,' without asking what the worker wanted. They just wanted to deploy cool technology.'

'We saw a great example at a utility company, they deployed expensive AR headsets, and said, 'this is going to help the factory speak to you.' The people they gave it to absolutely hated it.'

'It didn't help them do their job, they felt it was an imposition from outside. They dropped the headsets, left them at home, they did everything they could to make these things break, so they could go back to doing their job the old way. If the innovation team had bothered to talk to people and find out what problems they want to solve, it might have gone a lot better.'

Aarbakke's Rolf Thu also agreed. 'We firmly believe you should do digitalisation from the floor up. Machinists, welders, those are the guys creating value in the company. People who have all the craft, all the knowledge, you need to give them the ball to get into the goal. It's about listening, evolving and asking questions,' he said.

'We developed solutions. We said, 'this is the fancy one, do you want this?' and the workers don't need it. The worker wants, 'how can I improve, how can I work faster?''

'Let them learn from each other. If one guy on the night shift does something really smart, you can capture that with data and inform the other guys. I think learning is what it's all about.'

'The 3D digital twin is eye candy; it is something you can use to sell the company. It is really impressive to put some information on a digital twin and walk around. But it is not adding value. It is adding value maybe in terms of a sale.'

Give people what they need

Forrester's Mr Miller said he completely agreed with the point that a visual digital twin can be just 'eye candy.'

'We did some work at Forrester [to find] six key characteristics of a digital twin.

'One of those was the ability to report. That reporting might be pretty 3D eye candy if it is for a chief executive. If it is a field service engineer, it might be a text message change filter 7 on pump 23. That's all they need to know.'

'You're relying on intelligence in the system to be able to extract those insights and give them what they need to know to do their job.

'We see a lot of investment into building those contextual capabilities, understanding the asset, understanding the skills of the person. Is a text message good enough? Is a tag ID good enough?'

As an example of how not to give people what they need, Mr Miller cited a project from the UK Environmental Agency to build flood maps. People could look at the map to find out which areas would get flooded as water levels rose, and if their house would get flooded. 'That's not what users wanted. They wanted to type in their own address and have an answer 'flooding, yes or no.''

It is important to have 'the ability to solve an actual problem rather than get hung up on the power of a GIS.'

Rather than 'get too hung up on shiny machines and shiny algorithms,' we should be thinking, 'what's the shortest, simplest cheapest way to give the person the knowledge they need.'

A valuable digital twin

If a digital twin is defined as a tool which can provide all the data in the factory in an integrated format, it can be a strong source of value, Aarbakke's Mr Thu said.

With the data, 'you can develop use cases, you can slice it, you can see what's the value. You also have the possibility to be really quick on proving if a use case is valuable or not, you have historical data right away. If you improve, you save cost, that's basically what you want to do from the executive level.'

As an example, customers are asking about the CO2 emissions made in creating their product. With such a digital twin, 'we can now document and prove what CO2 level we have, down to the smallest process,' he said.

'I think that rather sooner than later there will be a disruptive change in manufacturing. Manufacturing is still in the stone age. We believe so strongly in this, we spun out this company to sell 'smart factory' as a service. What we have built is something that all the industry will need.'

'Future manufacturing will be different, integrating with customers on a level of data flowing between customer, manufacturer and end user.'

'Our customers want to see what we're doing. They want to see emissions transferred through [data] integration not through pdf.'

Associated Companies
» Cognite

CREATE A FREE MEMBERSHIP

To attend our free events, receive our newsletter, and receive the free colour Digital Energy Journal.