Iot Analytics

Unlocking The Value Of The Industrial Internet Of Things (IIoT) And Big Data In Manufacturing

Pinterest LinkedIn Tumblr

Hirotec is a tier-one Japanese automobile parts manufacturer, supplying components directly to makers such as GM, Ford and BMW. It also provides tools and expertise to other manufacturers wanting to create their own parts.

Low-code Application Development Company

In late 2015, the business began investigating the Internet of Things (IoT) “beyond the buzzword”, in terms of drawing up plans which had the real potential to bring about transformative change.

In light of reports that many businesses were (and still are) struggling to get real value from investment into Big Data-driven analytics and technology, a decision was made to invest initially in small-scale, short-term strategies where value could be seen quickly.

The plan, as instigated by IoT lead Justin Hester, was that these would work as proof to gain buy-in across the company for further, larger scale and more challenging projects.

“The promise of IoT is well understood in the industry,” Hester told me. “This idea that we can finally harness the data coming in from all of these different sources, whether they are machines, humans, parts – but I think the real challenge is the next step – how do I execute? That’s the challenge.”

Hester and the team at Hirotec’s answer was to start small. The thinking was that the reason behind the failure of many businesses to get to grips with implementing IoT analytics was an over-eagerness to “boil the whole ocean.”

“They say ‘we see tons of benefit from Big Data and we want to bring all of this in and analyse it’ and that’s true and sounds great – but data doesn’t collect itself and it doesn’t analyze itself.

“So, we recognized the need to create short sprint activities which were projects which would help us solve one of our internal challenges and also be scalable so we could implement them again but on a larger scale.”

Gathering real-time measurement

One of the early projects initially involved connecting and monitoring computer-controlled cutting devices at Hirotec’s North American tool building facility. Here, ultra-precise blades cut through metal and plastic and minor fluctuations in the performance or reliability of machines can have large knock-on effects.

The abundance of legacy technology in the manufacturing world was an initial problem – unlike in computing, manufacturing machines are expected to remain productive until they die of exhaustion, rather than being replaced with new models every two or three years when they start to get a little creaky.

“We had these machines on the shop floor and there was one from 1970, with no connective technology, and one from 1990 which had this really cool new tech that could send a message to your beeper.”

Cool in 1990 perhaps but pretty useless in 2016. Hirotec’s analytics partner, PTC, provided the solution to this with their Kepware platform. With the initial aim of finding out how a reliable, real-time measurement could be taken of all of the cutting machines across the workshop floor, Kepware acted as a “universal translator”, says Hester.

Data from Kepware is then transferred to the cloud-hosted ThingWorx Big Data and analytics platform – giving the real time access to performance data and insights that had been the goal.

“Not only can we see the status of the machine but we can see historically what the performance of each machine has been – is there a certain machine that sits idle at 10am every day? Now we know, and we can look at that and see how we can make it more productive.”

Visualizing and reporting

Following that success and increasing interest from the business leaders in the potential of IoT and analytics, lessons were applied to a larger project at the group’s headquarters in Hiroshima, Japan.

Having proven that data could be collected and analyzed in real-time, focus shifted on how it could be made actionable as efficiently as possible. For this project, a robotic inspection system (known as the Hirotec Inspection System – “You can tell it was named by an engineer!” says Hester) was chosen as the guinea pig.

Here the goal was to eliminate the “Tuesday morning meeting”. This phrase originates with an observation that traditionally in manufacturing, morning meetings would be held where the topic of discussion would be what had happened the previous day.

“We needed to bring real-time, contextualized data, that makes sense, and has supporting information around it – that is actionable – right away – they need it so they can do something about it on Monday, instead of waiting till Tuesday.”

The result – again implemented through Kepware and Thingworx, was a reporting system designed to communicate the insights that matter to the people in a position to take action. Output ranges from a two-light, green, yellow and red plant manager dashboard, down to department level, where output is in the form of a timeline, all the way to shop floor engineer level, where detailed diagnostics and statistics is available on individual parts and processes.

This is essential considering that each robotic inspection device is designed to measure 400 data points on each part – typically exhaust components – that passes along the production line. Presented with that data in raw form it would take a special type of human genius to draw out the vital insights – and even they would do it far more slowly than a computer.

Looking forward

This more or less brings us up to the present day, where Hirotec are currently in the third phase of expanding their analytics and data-driven transformation across the entire company.

The focus is currently on connecting an entire production line at one of its Japan plants, meaning that manufacturing of an entire complex auto part – specifically a door – will take place within the connected, industrial internet of things.

“Every robot, every heavy press, every inspection system – let’s take all of that data and visualize it for all of our team members – from line side to the c-suite. You can see how we’ve taken this model, scaled it up a little more, taken the lessons learned and applied them to continue to grow – but in short, manageable projects that always bring value to the business,” says Hester.

A little further down the line, Hirotec is planning on implementing augmented reality on the workshop floor. This will enable data about the performance of machinery to be relayed simply by looking at it while wearing headsets designed to overlay digital information and visualizations on the real-world we can see and touch.

“We’re able to take this phased approach because we’ve selected an ecosystem partner rather than a specific solution partner – we’re able to say ‘let’s give data in real-time to our employees and see what we can predict.” Hester tells me.

Hirotec’s approach to implementing IoT is a great example of how real value can be generated from analytics and connected technology when a connected and goal-focused strategy is rolled out in a manageable way. It’s never a good idea to run before you can walk and it’s important to remember that although much is said about the need to implement data-driven decision making throughout an organization in its entirety, it doesn’t all need to be done at the same time.

Starting small and focusing on building capabilities – whether that’s measuring, analyzing, visualizing or taking action – piece by piece, is a strategy that is less likely to lead to costly analytics failures, false-starts and misfires.

ThinkDataAnalytics is a data science and analytics online portal that provides the latest news and content on AI, Analytics, Big Data, Data Mining, Data Science, and Machine Learning. A team of experts with extensive experience in the field runs