The smart factory and connected operations: keys to success

Find out how smart factories can be delivered on a practical level. Where are we now, what are the barriers to developing successful smart factories, and how can they be overcome?

Rise of the machines

For anyone connected with the manufacturing industry, a revolution is afoot. The Industrial Internet of Things (IIoT) – a collective term for the introduction of connected sensors and devices, and centralised analytics engines, to industrial settings – is transforming warehouses, logistics centres and factory floors of all shapes and sizes. 

By forging closer links between design, manufacture, supply and demand, and maintenance, the IIoT is helping industrial organisations to be more flexible, more efficient and more integrated in their approaches – as we explored in our insight guide looking at how the IIoT can deliver smarter factories.

This insight guide moves the discussion along to look at how those smart factories can be delivered on a practical level. Where are we now, what are the barriers to developing successful smart factories, and how can they be overcome?

Where are we now?

According to a Capgemini survey of 1000 senior executives across a range of countries and key industrial sectors, just over three quarters of manufacturers either have an ongoing smart factory initiative or are working towards formulating one, while more than half have aligned $100 million or more towards smart factories. It seems clear, then, that we are past the point of discussion without action, or of early adopters being the exception rather than the norm. Nevertheless, the same survey revealed that only 14% of organisations are satisfied with their level of smart factory success. There is plenty of space for improvement, and innovation.

Where will those improvements come from? At this stage, we need to delve a little deeper into the technology that can constitute a smart factory.

The connected operations that underpin the IIoT are at the heart of any smart factory. Each needs a mechanism for capturing data, consolidating and analysing it, and feeding that data intelligently back into operations. The required technology is a machine-to-machine communication system, comprised of wireless communication channels such as radio-frequency identification (RFID) or near field communications (NFC), and a centralised analytics engine or platform. That feedback loop mechanism, through which tangible business decisions are made on the basis of previously untapped data, and in many cases entirely automatically, is the heart of what makes the smart factory smart.

 

ThingWorx_Studio

The feedback loop is augmented with a wide range of additional technologies and processes, including:

  • Big data analytics: If the smart factory, at its core, is all about harnessing previously ignored data, then clearly smart factories must handle a far vaster quantity of information than any previous industrial setup. In turn, this requires analytics engines and platforms with the processing power to handle big datasets – that is, to consolidate them, filter them, analyse them and translate them into understandable business information.
  • Automation and robotics: Many processes within modern factories are already partially or even entirely automated, thanks to the role of robotics in putting together complex components. However, the smart factory extends the role of automation out into data processing and takin action from that data, and the role of robotics out into robots that respond to their unique environments and are extremely adaptive. Such intelligent robotics can expand out of the production
  • Augmented reality (AR): AR technology blends the physical and digital worlds via overlays of imagery, video, text and other content, using anything from regular smartphones and tablets to highly specialist cameras, motion sensors and depth sensors. In turn, this technology can provide on-the-spot training or instructions to staff, driving closer and more precise partnerships between personnel and the machines they operate.
  • Artificial intelligence (AI) and machine learning: These technologies enable elements of the smart factory to enhance and optimise themselves, rather than waiting for human input. This can then drive a broad range of simple yet effective benefits, from less downtime and fewer outages to the removal of process bottlenecks.
  • Additive manufacturing: More often called 3D printing, additive manufacturing enables industrial organisations to deliver highly personalised solutions according to customers’ precise requirements, and to experiment more flexibly with different solutions and products.

The smart factories of today – and tomorrow – incorporate a range of these technologies, adapted and tailored to the precise setting in question. There is no ‘one size fits all’ smart factory.

Delivering a smart factory: the challenges

The list challenges when it comes to developing a smart factory is long and diverse, but some of the major areas to think about are as follows.

Design

A smart factory cannot be bought off the shelf. It needs to be carefully designed and tailored to the needs of the organisation in question, even when it is based around non-proprietary technology. This requires a logical and pragmatic understanding of the organisation’s needs and goals, and careful project management throughout.

Cost

There’s no getting around it – creating a smart factory costs money. Deploying brand-new connected hardware on the factory floor clearly requires an upfront investment, but so too does retrofitting existing equipment with connected sensors. On an ongoing basis, whilst smart factories should ultimately drive cost efficiencies in terms of streamlined processes and shifts to models like predictive maintenance, they still involve different financial models to legacy factories. Measuring those costs against the benefits realised by the smart factory, and therefore evaluating the success of a smart factory project, can also be challenging.

Skills

The IIoT marks a radical shift in how industrial hardware is managed, operated and maintained – and this means that in-depth training is often necessary to enable staff to adapt to the new setup.  Maintenance and engineering staff, for example, are required to move away from old processes of regular manual checks, and plan their schedules based on data automatically generated by the factory floor. Production and assembly staff may be required to start using augmented reality technologies. Upskilling existing staff members to use this new technology requires additional investment in terms of time and money.

Security

The smart factory is, by nature, a connected factory – and this means that protecting it against digital vandalism and deliberate infiltration has to become a major priority. Cyberattacks on the smart factory can have a devastating impact in terms of day-to-day operations and long-term protection of assets, so it is vital that organisations developing smart factories can both defend against cyberattacks and rapidly identify, isolate and mitigate them if and when they do happen.

Governance and compliance

Much of the smart factory’s intelligence – its ability to learn from itself and optimise processes without human interaction – fundamentally transforms the ways in which organisations manage their internal processes and achieve and demonstrate regulatory compliance. New processes and audit trails may need to be developed. 

 

Keys to success

How, then, can industrial and manufacturing organisations best meet these challenges and develop smart factories of the future?

Just as there is no ‘one-size-fits-all’ technology solution for building a smart factory, so there is no single path to success for creating one. However, there are some principles to follow.

Project management principles: aims and objectives

Too many industrial organisations still approach the smart factory as in inevitable next stage in the evolution of their industry, rather than a proactive IT or technology strategy. However, thinking of the smart factory like any other business transformation project encourages a starting point of what is our business trying to achieve?

From there, it is far easier to pin down what kind of data will ideally be captured by the smart factory, and what kind of insights and actions that data could drive. And since the smart factory is, at its core, all about capturing and harnessing data, this is the best way of ensuring a smart factory that is closely aligned with business goals, and a smart factory whose success can be easily evaluated.

Five key stages

Once you have confirmed your success criteria, all operational smart factories, no matter what their precise technological setup, scale and scope, should ultimately follow five practical steps:

  1. Implant sensors to capture data, whether through retrofitting existing equipment or installing all-new hardware.
  2. Create an architecture to consolidate and analyse that information, with coping with the different protocols between all equipment using Kepware and comprehensive network security throughout.
  3. Deploy an Industrial Innovation Platform such our Thingworx platform, which can process big data from the equipment and business systems and transform it into meaningful, actionable intelligence.
  4. Put in place the mechanisms and processes to use those insights to drive tangible business actions.
  5. Augment and improve the smart factory, with processes like AI and machine learning so that your IoT investment is maximised on an ongoing basis.

However, those five key steps do not need to be applied across your entire organisation all in one go. In fact, from a cost-efficiency and ease of project management perspective, it is normally sensible to work on making disparate parts of the factory ‘smart’ at different times.

Work from the inside out

Your people will always be key to making the smart factory an ongoing success. You need to spend time explaining your objectives and listening to and engaging them on the journey. The project cannot get off the ground without technology alone. As such, the most logical approach to developing a smart factory begins with people and the business case and then moves to the equipment. Why are you doing this? How will this affect the team? Who will be responsible for this new approach after you have implemented? Will it change the way your people are rewarded? Do they have the right skillsets?  How will the human/machine collaboration work in your smart factory?

Only then do you get onto the technology. Where do you need to retrofit existing hardware, and where will you install brand new equipment? How do you cope with the many proprietary protocols? Which centralised analytics engine is most appropriate for your goals – today and into the future? How can you maximise the useful lifespan of all the hardware on your factory floor, via a predictive maintenance programme? . And from there, you can extend your smart factory out beyond your manufacturing premises, to consider your supply chain and logistics.

Setting up a smart factory doesn’t need to be complicated, and it doesn’t need to involve ripping out your existing setup and starting again. Much better to “Wrap and Extend.” A logical approach can create a dynamic, intelligent and proactive factory floor, positioning you strategically for the future.