You may not see the immediate benefits of the Internet of Things (IoT), but chances are someone using your equipment or machine will. Heiko Luckhaupt, Industry Sector Marketing Manager, RS Components, suggests a simplified step-by-step approach to clear up some of the questions.
The Internet of Things is nowhere near its critical mass yet, as new concepts and ideas that exploit the connectivity it offers are being developed all the time. Indeed, it is hard to read the technical and even national press without stumbling across new and interesting innovations that are leveraging the Big Data being generated in just about every walk of life.
Industrial take-up is also gaining pace, with the most obvious manifestation of the IoT – in an industrial sense – being the concept of the Smart Factory, where even the dumbest component can be interrogated by higher-level systems using communication (Ethernet) capable middleware.
So, what do you need to do to become part of this new industrial revolution? This simple nine-step plan will provide the basis for some questions you can ask yourself to see if it will work for you, or indeed your customers.
Why do you need to consider it?
You may not need it yourself at this point in time, but chances are that somewhere in your value stream there is a demand for greater data capabilities. Top level customers, including many of the worlds leading end users, are already operating in data-rich environments and they are exploiting this data to streamline and enhance their operations – not just in production but also in just about every element of their operations, from incoming logistics all the way through to final sales. If you don’t give them the capability to easily tie your technology or products into their systems, then you could miss out!
Identify the opportunities
Big data means better decisions can be made in real time in order to have the biggest immediate impact. The smart factory concept is based around the principles of knowing what is going on, how well it is all working and how it can be improved – both in terms of performance and maintenance.
It’s not always immediately obvious from the raw data, but there are some very clever analysis systems out there that can make sense of this tidal wave of operational data and then provide the most pertinent elements in an easy-to-digest form. Couple this to advances in artificial intelligence and it is a huge recipe for success.
Identify crucial data points
What parts of your product or machine can divulge useful, actionable data? Do you use any sensors or feedback devices? Can speed changes make a huge difference in throughput? Does line balancing affect output? If you have any data that represents operational performance or metrics, chances are it can be exploited. Even a simple sensor to measure and feedback bearing health on a critical piece of machinery can make a huge difference. Imagine you are in a factory that relies on air tools; if your compressor was to fail the impact is almost too huge to comprehend. With the proper data paths in place you would be able to identify failure – or less than optimal running – much earlier, and with historical data you can even identify trends relating to failure patterns, based on shifts, time of year and loads.
What network connectivity is involved?
If you are using lower level fieldbus protocols, you may need to invest in smarter controllers that offer Ethernet, or higher-level communications capabilities. Costs are not prohibitive, and even slightly more powerful controllers could actually improve your machine’s operation. Wireless is another option. As long as you can get your data in the hands of a device or software package that has the ability to feed enterprise systems, then you are on the right path. Many handheld devices exist that operate using RFID or Bluetooth, but even these can eventually liaise and share data with Ethernet enabled systems.
Identify hardware needed to extract and share this data
As discussed above, a simple controller update may be all that is required. You can also look to exploit intermediary devices that collect data and can then share it over an Ethernet network. Even a basic PC computer shared by multiple cells or machines can act as the ‘middle man’, using software specifically designed to collect and share operational data.
Developments in Ethernet technology and the commensurate drop in prices have led to smaller commodity devices being big-network compatible. Process sensors are a great example. Many exist with Ethernet connectivity, and for the hard-to-reach/-wire and remote areas wireless technology has started to see massive uptake. All that is needed is some form of collector – often seen in the form of ruggedized industrial wireless modems/routers. Setting up a communication cloud in a factory or remote plant is now an everyday occurrence.
Identify how you are going to collect, decipher and present this data
It may be, as a machine builder, that you don’t need to go up to this level, although presentation of diagnostic information is an incredibly powerful selling point. But if you are in a factory that is collecting data from multiple sources, you need to develop a plan on how you will collect, collate, decipher and disburse the data in a form that will allow you to make these all important decisions. There are many analytical software packages that allow you to create custom dashboards for visualisation operational parameters, indeed most of the leading PLC vendors have just such an offering and are in a great position to be able to help you with this data gathering. These analysis systems are often highly customisable too, meaning that you can create a common or branded look and feel across all elements of your production plant.
What sort of security is needed?
Multiple connected devices ultimately equates to a greater number of threat vectors for unwarranted access. This is a widely recognised issue and you only need to look at www.Shodan.io to see the worrying number of unsecured IoT devices out there. However, the major vendors have once again got your back, so to speak.
Security is up there with performance and safety in their eyes and there are numerous industry proven security solutions – using dedicated hardware and software based “no man’s lands” that act as a buffer between IT network and the shop floor – that can keep out even the most determined ne’er-do-well.
Many large end users are beginning to converge their IT and manufacturing networks as the platforms and architectures evolve to become very similar; and leading IT suppliers have already identified the issues, threats and opportunities and have system and solutions in place to cater for these demands.
So, what can you do with all this data?
You may not be the primary recipient for this data, but somewhere down the line there will be someone who will be glad to see it. It may even be for historic purposes. Data you collect now can be used for trending and monitoring purposes in the months ahead, either by you or by the end user of your equipment. The primary idea is that if you have access to data everything at some stage, parts of it are going to prove useful for someone – if not you, then your customer or their customer.
Identify what sort of decision can be made
This is the kicker. It may not be immediately obvious what can be achieved with all this data. Historic data could start to indicate trends that were not apparent in the short term. Short term data can be used immediately to make even minor but highly impactful modifications to production parameters and, if we jump forwards a few years, artificial intelligence will soon be making its presence felt across multiple industries for massive optimisation opportunities. And what will it rely upon? You’ve got it. Data!
