Jan Hemper

Blog by Jan Hemper - 25 January 2018

Selection, deployment and utilisation of appropriate technology enhances almost everything: improving quality, reducing costs and generating new revenue streams. As Technical Director of InVMA, I analyse functionality and performance requirements to architect and to advise on systems that provide real, useful and tangible business benefits through the use of new and emerging technologies.

Five steps to making the smart factory a reality

 

‘Responsive, adaptive, connected manufacturing.’ That’s how Deloitte characterises the smart factory – and those adjectives underline exactly how beneficial moving to such a setup can be. Smart factories involve embedding IoT sensors throughout the factory floor and using them to capture and analyse data pertaining to machine performance or manufacturing process. In turn, these insights are used to deploy a more proactive, strategic approach to hardware maintenance, to enhance production and sometimes even to develop whole new products and processes.

All of which sounds very positive. But, on a practical level, how do you go about actually setting up a smart factory? If you’re starting off with a legacy, unconnected manufacturing operation, how can you deploy connected technologies without tearing up what you have and starting again? How can you make the smart factory a reality?

Here are the five key steps you need to consider:

1. Capture existing sensor data and implant new sensors

Purpose-built, IoT-enabled equipment for the smart factory does already exist, but most organisations already have a significant amount of sensor data that is simply not captured or captured in isolation. Transitioning to a smart factory will mean consolidating the capture of existing data sources and augmenting that with new intelligent sensors where required. These new and existing sensors need to be able to capture simple data which may vary from  temperature to oil levels, information pertaining to wear and tear or the speed or number of revolutions a machine has completed.

2. Create an architecture

It’s all very well going to the trouble of collecting all that data, but it’s useless unless it can be consolidated and contextualised. As such, the next step is to think about the connectivity between all these sensors – and between the sensors and some kind of centralised analytics engine. This means designing a network architecture, and potentially joining up a number of existing disparate networks – and, in turn, standardising network protocols between different pieces of equipment. Network security is a vital consideration at this stage too – all data should be fully encrypted in transit as well as at rest.

3. Deploy an analytics engine

Next you need to consider how you can turn that wealth of newly collected data into real business insights – and this requires a powerful IoT analytics engine, which can process big data and transform it into meaningful, actionable intelligence presented to diverse teams in meaningful ways. Our Thingworx platform is a prime example.

 Crucial factors to bear in mind when selecting such a platform or engine including the range of different devices and systems it is able to integrate with, the speed and ease with which the applications can be developed and the kind of interface it can provide to a varying user base. A platform that can only be used by highly technical members of your team is restrictive – far better to choose a tool that makes sense to all members of staff.

4. Use insights to drive action

This is actually a non-technical step in setting up a smart factory – and yet it’s a step that is frequently forgotten. Carefully implementing all of the above steps is wasted if the data they collect and the insights they generate are not then fed back into business processes – it’s this feedback loop that makes the smart factory, well – smart. Typical actions within smart factories include adjusting maintenance schedules so that costly manual checks are no longer needed and maintenance takes place at the most efficient time, and rearranging production processes to remove bottlenecks and optimise manufacturing rates.

5. Augment and improve

The above four steps should help you to create an ongoing smart factory, which continually collects, analyses and actions data from across the factory floor. However, this process can be made even more dynamic by harnessing processes and tools like artificial intelligence (AI) and machine learning, so as to optimise your smart factory setup as you move forward. Thingworx, for example, integrates machine learning and augmented reality solutions so that your IoT investment is maximised on an ongoing basis.

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. With these five steps, you can create a dynamic, intelligent and proactive factory floor, positioning you strategically for the future.

Topics: IoT, plant floor machinery, machinery, smart factory

Industrial Internet of things

Comment on my blog