For companies in the manufacturing and logistics sectors, the new era of instant demands can be better met through more use of Industrial Internet of Things (IoT).
The Industrial Internet of Things involves the use of IoT technologies in manufacturing processes and across supply chains. Alongside data from devices and sensors, Industrial IoT strategies should incorporate machine learning and big data technology, harnessing that combination of existing sensor data, machine to machine (M2M) communication and automation technologies to provide more insight back to the business.
According to Industrial Analytics 2016/2017, 69% of decision makers see industrial analytics as an essential way of achieving goals and therefore has a considerate impact on a company’s ecosystem.
The potential of technology here is vast, with new advancements being developed all the time and at great speed. Companies must therefore consider how best to adopt Industrial IoT as part of an ongoing plan to benefit the business. In turn, this level of insight can contribute to future business decisions and levels of success.
What advantages does industrial IoT bring?
Manufacturing enterprises tend to have large volumes of industrial equipment, all of which needs maintaining. For existing deployments, Industrial IoT enables improvements in decisions around manufacturing processes based on availability of more accurate data. It can also be used to improve production quality and uptime, as the data gathered from devices and sensors on the network enables real-time and predictive maintenance across the estate.
The main idea behind Industrial IoT is that of making machines smarter and more efficient than humans at making decisions. This does rely on accurately, consistently capturing and communicating data. Companies within the market are now developing leading edge sensors which make highly accurate measurements. This data can be coupled with real-time analytics that can explain how well machines are performing.
Using machine learning, the system can be trained to spot potential patterns that would indicate a future failure; if the results are concerning they can can be investigated immediately. This information would previously take weeks to discover and rely on the availability of skilled professionals at every site; now, the use of real-time data can help those with the right skills monitor more machines in multiple places, making decisions on maintenance faster. In turn, the efficiency of manufacturing can speed up considerably.
>See also: The rise of IoT in industrial organisations
Alongside predictive maintenance, machine learning heavily lends itself to Industrial IoT around how to make improvements in performance. To make this work, companies need to make sense of all the data available, quantify it and provide insight into how best to proceed with their manufacturing processes given the information received. Overtime, machine learning can be used to demonstrate how to improve performance and provide faster results for the same level of investment.
In manufacturing, time is money. With the implementation of Industrial IoT strategies – from individual sensors through to the analytics and automation available – manufacturers can make better business decisions around overall efficiency and costs. Industrial IoT strategies also hold great potential for improvements around sustainable and green manufacturing practices, and supply chain traceability.
With the growth of companies in the “circular economy”, where all elements from products through to packaging should be fully re-useable and recycled across the value chain, Industrial IoT implementations will be essential to tracking results.
Why security is crucial for Industrial IoT: The threats businesses should be aware of
Alongside implementing Industrial IoT systems to meet business needs, companies should be aware of the security implications too. Each Industrial IoT device that joins a network becomes a potential point of entry for attackers, so the need for a secure network couldn’t be more critical as the amount of devices used continues to increase.
Attackers are getting more creative in how they try to access networks – security company Darktrace recently discovered that one malware group had tried to enter a casino network via an Internet-connect fish tank.
Anything with an IP connection should therefore be secured before being allowed on to the network. Each device should also have an upgrade and patching path, so that any flaws can be fixed.
When it comes to information technology versus operational technology, there is currently a lack of established best practices. However, it is worth using some of the standard operating procedures for IT security around Industrial IoT management systems as this approach provides a head start to securing networks. These management tools should be treated as though they are attack vectors in order to keep networks secure.
There are eight elements of security required at the edge of an Industrial IoT system, to ensure that the whole system remains protected against attack:
1. Authentication – anyone that requires access to a network needs to be authorised. Implementing an authentication strategy with role-based access control and auditing to track who is allowed to join the network, and when those accounts are used, is therefore an essential part of your IT infrastructure strategy.
2. Encryption – to have control over data being shared from devices back to the central application, encryption is required. However, this encryption should not affect the speed at which the data is processed adversely.
3. Transmission – the transport mechanism for getting data from devices from one device to another, or to central systems, should be considered. Will devices be on a wired company network, or will they be transmitting data over wireless networks? For logistics and supply chain companies, use of telecoms networks to track vehicles may be necessary. The method of transport whether wired or wireless needs to be secure and flexible to upgrade to latest communication protocols.
>See also: The Internet of Things: Success or bust?
4. Tamper protection – this step ensures that equipment is monitored and checked to see if it has been tampered with. This can be a combination of mechanical, hardware and software tamper detection.
5. Data storage – all data storage must be secure. Data is often being stored at the edge and can hold critical system data like operating software. Additionally, over the life cycle of equipment, memory expansion is very likely and needs to be implemented with no loss of security integrity. Similarly, the central set of data also needs protection, as it forms a critical business asset.
6. Over the air software upgrade – the lifespan of industrial equipment varies and often has to be updated. An Industrial IoT system needs to be able to do over-the-air updates in a secure way.
7. Segmentation – flat network structures are most commonly used in production environments and any flaw in a device can spread and effect the whole network. Segmentation is therefore used in operational technologies to keep the attack surfaces small and minimise this risk.
8. Multi-layering – Operational technologies can’t be easily patched and so Industrial IoT often sits above the operational technologies network and is operated in a separate network layer; this then communicates with data aggregation and an analytics layer to secure a network. As additional sensors and gateways communicating to IT infrastructures outside the network become threats, security in layers such as this is used to protect the production environments.
As the breadth and depth of technology available increases, ensuring a secure Industrial IoT network is imperative.
Sourced by David Fearne, technical director at Arrow ECS