- By Louis-Philippe Lamoureux
- October 06, 2021
The five core challenges process engineering businesses face–and how an Intelligent Edge approach is key to more successful edge deployments.
The industrial sector is undergoing a rapid evolution in its digital transformation journey–from a pure Industrial Internet of Things (IIoT) focused predominantly on connecting assets, to an Artificial Intelligence of Things (AIoT) that puts AI-driven industrial intelligence at the heart of operations. This transformation from IIoT to AIoT has exposed the risks and challenges that process engineering businesses face by displacing their data beyond the edge and outside the plant–and the need to adopt new edge technologies capable of working in concert with cloud workloads to mitigate those risks.
Enter the Intelligent Edge. The Industrial Intelligent Edge is the integration, mobility and analysis of industrial data with edge technologies working in concert with cloud workloads, to enable the deployment and management of AI-enabled solutions both at scale and at the edge. In this way, the Intelligent Edge offers a way forward for industrial organizations to reduce latency, optimize costs, and mitigate security risks, while helping the business operate with improved profitability, productivity and sustainability.
The industrial sector needs to rethink its approach to displacing data beyond the edge and outside the plant. Here are five core challenges and risks that processing businesses face by sticking to this old model–and how an Industrial Intelligent Edge approach can resolve them and enable more successful and secure edge deployments.
Delayed responses in sending data to the cloud cost industrial organizations valuable upload and download times. Depending on where your operations are located and the size of the data being sent from a device to a centralized server, you could be looking at transfer times ranging from two minutes for 1GB of data to four months for 100TB. Even taking the low end of those estimates, ask yourself: is two minutes really an acceptable wait time when performing critical controls over your industrial assets? For monitoring or supervision? Even a two-minute lag in latency between the cloud and back can have negative consequences on infrastructure and operations. Slow response times make it impossible for industrial organizations to have literal real-time insights and analysis into their operations.
As enterprises shift more confidential information to the cloud, their cyberattack surfaces are only getting wider. Consider how data breaches targeting cloud-based infrastructure jumped by 50% from 2018 to 2019. This year’s Colonial Pipeline ransomware attack was a particularly high-profile example of the devastating consequences of a successful cyberattack aimed at industrial assets.
Moving data out of the plant increases the number of potential cyberattack vectors. It’s not just outsider threats to consider, either. When part of your IT infrastructure is outsourced, attack sources can range from malicious insider threats to simple misconfigurations that result in breaches. Trying to ensure data security in this kind of dynamic environment is extremely challenging.
3. Digital sovereignty and cyber regulatory compliance
When industrial organizations displace so much of their data beyond the edge, they’re creating a disproportionate balance between the data they have sovereignty over and the data that is out in the wild. How much can we trust those who manage our cloud servers storing that data? There’s too much control being ceded to too few places. Industrial data outside the plant changes hands across so many third parties, before ultimately being controlled by a small number of large tech companies and cloud providers. How much digital sovereignty do you have over your organization’s IT and OT infrastructure in this status quo?
When pieces of your IT infrastructure are outside of your scope, you need to be able to trust who owns those pieces. This isn’t just an issue of data management; it’s a compliance issue, too. If your data ends up in the wrong hands, how liable are you for that? Data movements from the plant to the cloud may violate certain regulatory compliance standards along the way, carrying new liabilities and potential penalties for the business.
4. Cloud costs
Despite a current annual public spend of $500 billion, the cloud is still in its early days. Industrial organizations’ cloud needs are only going to continue to grow rapidly, and in tandem with that will be an equally high growth in cloud-centric implementation costs. Network costs will rise significantly as industrial organizations deploy more cloud-connected devices or digital twins.
The cloud is and will remain a business-critical function for any industrial organization going forward. The cloud helps cultivate innovation by freeing up company resources that can instead be focused on new products and company growth. But, the rising costs of cloud implementations point to the need for more organizations to repatriate some of their IT infrastructure back to an on-premise or hybrid environment, in order to alleviate cloud costs.
5. Domain knowledge retention
There is a generational churn occurring in today’s industrial workforce, with employees who held their jobs at the same plants for years or decades now retiring. These veterans are being replaced with younger employees who do not have their predecessors’ level of historic, operational expertise. As a result there is a growing brain drain of domain knowledge that is exacerbated by a high attrition rate among younger workers. Retaining the domain expertise of both retiring veterans and younger generations moving on to new jobs is business-critical to an organization’s IT/OT management and competitive advantages.
Resolving industrial data challenges through an Industrial Intelligent Edge approach
With an Intelligent Edge approach, there’s no data to transfer from the plant to the cloud, enabling a faster actionable response with a distributed processing mechanism. Reducing your cyber attack surface and keeping data behind a firewall provides cybersecurity teams with more oversight on potential vulnerabilities. Repatriating workloads back on premise through an Intelligent Edge approach allows for more control of your digital destiny – the data, hardware, and software you create remain within your digital sovereignty, not a cloud provider’s. Moving this data back on-premise provides greater transparency and confidence that you’re operating in line with the required regulatory compliance standards. At the same time, this repatriation of IT infrastructure from the cloud to the plant naturally reduces cloud costs. Finally, an Intelligent Edge approach alleviates the challenge of domain knowledge retention by automating parts of the machine learning pipeline--from training, to deployment, to retraining -- ensuring that organizations are able to retain domain knowledge and operational expertise even as individual employees move on.
While the cloud remains a key component of industrial IT and data strategies, shifting most of your organization’s workloads outside of the plant and into the cloud can be a dangerous crux, exacerbating the risks and challenges listed above. Taking an Intelligent Edge approach to data provides a best-of-both-worlds solution: leveraging the value of cloud computing with the security, latency, cost and data management benefits of keeping data in the plant and at the edge.
Did you enjoy this great article?
Check out our free e-newsletters to read more great articles..Subscribe