The IIoT wave takes manufacturers by storm
Syntax is a multi-cloud, multi-ERP managed cloud provider for mission critical applications. The company supports small- and medium-sized businesses navigating their cloud and ERP journeys. With regards to IIoT, the company address this through various approaches, but can basically be grouped under two areas: Smart Factory and Smart Product.
Smart Factory has the customer’s shop floor digitisation at its core with the overall target to optimise OEE, reduce production cost and simply optimise the manufacturing related workflows. This scenario can start from a Manufacturing Execution System (MES) angle and reach from green-field implementations and migrations to native cloud-based solutions, and can include topics like the replacement of shop floor computers by handheld devices or the retrofit of machines to make them smart.
The Smart Product scenario focuses on the product manufactured in order to either move to a digital business model, i.e. selling services instead of products, or improve the service offering around the products, especially in the sense of predictive maintenance and predictive quality. Syntax holds a long-standing history in the manufacturing industry and hence vast experience, not only in acquiring and processing machine data to generate insight and transparency, but more specifically to combine it with ERP process data to generate actions and feed the data back into the production or logistic processes to automatically optimise the whole production chain.
Syntax pairs its experience with turnkey solutions covering between 80-90% of customer requirements, leaving room for customisation and individualisation to meet the customer requests. This allows a fast turnaround of projects into customer success.
I asked Jens Beck, Director of IIoT, Analytics and Innovative Cloud Services at Syntax about the growth of Industrial IIoT, he said: “In my experience, I would not necessarily call growth slow. A lot of companies were actually very early to enter the game of industrial IoT, especially the automotive industry, since the promises of optimised OEE, reduced cost, faster production were very promising. Why the topic did not really kickstart as expected is a combination of three aspects:
- Digitisation is not simply plugging a ‘smart solution’ into your devices and then, as if by magic, you get all the positive outcomes, insights, and actions.
- Most companies have a quite heterogeneous shop floor with some of the machines not even being ready for gathering data.
- Most companies’ maturity state with regards to data governance and the importance of data was not where it is today.
- At the start of the Industry 4.0 wave the offerings from a technology standpoint, as well as the availability of experts, was very limited, and the upfront investment cost was rather high.
This also led to a lot of early adopters collecting first negative experience and the reasons above are still fairly common as to why IoT projects fail today. But all these reasons have changed:
- Customers realise that digitisation consists of various steps and while some of them are low hanging fruit, automation with the help of AI, requires human input, data, time and learning.
- Sensors, industrial gateways, mobile devices, or other edge components to retrofit non-smart machines are broadly available at affordable prices.
- Especially with the micro-services from hyper scalers such as AWS or Microsoft Azure, solutions can be implemented with a low initial upfront investment.
- Skills are available and there are plenty of solution providers.
Therefore, industries are fighting back by starting their journey with overseeable first steps to address the disruption IIoT could bring them.
What are some of the challenges associated with connecting industrial assets?
Beck explained: “Let’s start with the fact that we still find plenty of machines on the shop floor that do their job, but are not smart at all, some of them not even having a PLC. If you try and overcome this challenge with retrofitting, you find yourself in the situation that even your smart machines do not speak one common protocol. But this can be overcome by bringing the intelligence of translation to the edge via gateways, industrial computers or by smartening up your device. These are the most common technological challenges, but are all resolvable. The remaining challenges are more people driven.
“IIoT connects OT with IT. A connection many companies avoided on purpose in the past, for performance reasons or simply for security constraints, and with our solutions we even move data processing and analytics into the cloud. You might imagine that this even raises more concerns at times. You can address these topics with separated networks - virtually or physically, hardened edge devices, end to end encryption, trusted communication channels, two factor authentications via certificates and similar measures.
“Performance or better latency issues can be addressed with replicating access points, developing bundled communication patterns depending on the criticality of data, to name a few. A temporary non-availability of services needs to be architected into the solution, i.e., you need clear RPO, and RTO concepts supported by contracted SLAs to avoid data loss.
“Now, if you’ve followed me this far you understand that for static assets like stamps or grinders this does not sound too hard, since those devices are located on protected grounds in your shop floor. But if you remember my second scenario I mentioned earlier, the Smart Product, those might be part of a shop floor machine, but as well it could be something like a drilling machine or an electrical engine. So, you need to revisit the security and especially the availability architecture, and you need to understand where your product will end up designing the optimal solution scenario.”
How can companies seek expert support to manage IoT systems with minimal down-time?
“If you work with an IoT provider, ask straight away for RPO and RTO concepts and SLAs, ask about their operation offerings. A provider like Syntax offering you all services from the edge to the user out of one hand certainly has the answers you are wishing for, since our solution architecture answers the minimal down-time approach. We will ensure that critical components are highly available, and failure tolerant. We will guarantee SLAs on their solution and provide a service desk, a DevOps, and an application support team to ensure utmost availability. We will have tool sets available so you can monitor those promises.”
Research has found that most workers now prefer a hybrid working environment, Syntax supports this natively by ensuring all its components are manageable remotely, so except for the initial setup there is no need to physically interact with its components. The data collected is made available via the browser or apps to optimise the mobile experience. This reaches from simple condition analytics, via alerting, to confirmation of proposed actions for execution, e.g., you get an alert that the quality of your product is deviating due to an increase in vibration and the proposed action is to lower the speed of a conveyor. You review this proposal on your mobile phone, approve it and the speed is adjusted. Of course, with time such proposals can also be automatically executed if the trust is built. This way on-premises operations can be reduced and support further hybrid working scenarios.
“We’ve already seen with many customers that the physical presence of a worker at a machine to supervise it can be reduced significantly, as well as the miles walked by a worker. If we agree on the understanding that remote work means a worker does not stand aside a machine ‘baby-sitting’ it, I would subscribe that this becomes the new normal.
“Will this understanding also mean that in certain industries, production runs will be autonomous or nearly autonomous? Yes, I do believe so. Imagine a mill. There would be smart silos that store the raw material, the mill itself will be fully equipped with sensors, the machine learning models trained to react on varying parameters and not raising alarms but will send notifications way in advance for maintenance. At the end smart silos store the flour. I can well imagine that this is operated with very limited physical presence.” explained Beck.
When companies want to adopt a strategy to manage, analyse and secure their IIoT devices for ongoing remote work, they will vary from company to company. There is no one size fits all, since the demands and the cost benefit analysis are different from case to case.
For a first proof of concept, companies might want to skip the security concerns and the requirements for remote manageability of the IoT architecture. But once you’ve had your first success getting insights, you’ve got to trust in the approach. Companies should revisit the strategy. Before starting a massive roll-out or implementation programme, the strategy needs to be clear, as does the criteria for the different involved components from edge to user/action.
Beck explained: “The way we support our customers is two-fold. We offer a quick starter package at a manageable investment to get trust. On the other hand, we have developed a standardised questionnaire in order to get the bigger picture and to start creating awareness with our customers. The way we select components to be added to our turnkey solutions is via a checklist to ensure the base criteria for at least one of our core use cases is addressed - the more the better.”
How has COVID affected the IoT world and what does the future hold?
The pandemic itself has of course been terrible beyond words. However, for IoT and especially IIoT, it has been a tremendous accelerator. Lots of companies faced a serious impact on their supply chain and yet failed on having transparency on their production processes to fully be able to understand the criticality of this impact. Raw materials became rare, so avoiding scrap became more important than ever. Once the supply chain was up and running again, the shop floors needed to run at the highest possible to satisfy the appetite of the next tier.
“For many companies, COVID has re-emphasised the importance of an optimal production in the sense of higher OEE by reducing maintenance, and by simply adding shopfloor transparency at any time and from anywhere. So apart from that couple of months where COVID slowed down the economy worldwide I would say COVID has been an accelerator for IoT.”
2021 will be a prosperous year for IIoT and so will be the years to come. The demand is high, the potential is huge. Beck explained: “The challenges will be that on the one hand, the demand from customers is higher than the delivery capacity by solution providers - having said that, we need to differentiate what the objective of the customer is. Just getting a visualisation on a machine’s condition is in many cases 100% turnkey, but if it comes to more sophisticated, integrated approaches between machines or between machine and process data, this is where the delivery capacity will be critical to meet all customer demand.
“On the other hand, IoT solutions sprout like mushrooms from the ground, and it is hard for companies to keep an overview. The IoT world is very rapidly changing and so you need to keep your landmarks in sight and stay open to integrate what is useful for you, change what can be optimised and be able to simply adopt.”