The debate between cloud and edge computing strategies remains a point of contention for many controls engineers in the packaging industry. However, most agree that smart factories in an Industry 4.0 context must efficiently collect, visualize and analyze data from machines and production lines to enhance equipment performance and production processes. Using advanced analytics algorithms, companies can sift through this mass of information, or Big Data, to identify areas for improvement.

To some, edge computing devices may seem to create an unnecessary step when all data can simply be handled in the cloud. Microsoft Azure, Amazon Web Services (AWS) and other cloud platforms offer limitless space for this purpose. Moreover, MQTT encryption and data security built into the OPC UA protocol ensure that all data remain secure while in transport. When it comes to analytics and simple data management, however, edge computing presents important advantages to closely monitor equipment health and maximize uptime in production.

Because of the massive amount of data that modern machines can produce, bandwidth can severely limit cloud computing or push costs to unacceptable levels. New software solutions for PC-based controllers, such as TwinCAT Analytics from Beckhoff, allow controls engineers to leverage advanced algorithms locally in addition to data pre-processing and compression. As a result, a key advance in analytical information is the concept to process data on the edge first, which enables individual packaging machines and lines to identify inefficiencies on their own and make improvements before using the cloud for further analysis across the enterprise.

Bandwidth Burdens when Streaming Machine Data

Performing Big Data analytics in the cloud exclusively often proves expensive in terms of storage space. However, the more difficult proposition is first getting your data there. Managing bandwidth can create a serious issue for factories, since the average Ethernet connection speed across the globe is 7.2 Mbps, according to the most recent connectivity report from Akamai.

When one machine sends data to the cloud, much less multiple machines, little to no bandwidth is available for the rest of the operation. Two use cases published in a 2017 article by Kloepfer, Koch, Bartel and Friedmann illustrate this point. In the first, the structural dynamics of wind turbines using 50 sensors at a 100-hertz sampling rate required 2.8 Mbps bandwidth for standard JavaScript Object Notation (JSON) to stream all data to the cloud. The second case, condition monitoring of assets in intralogistics, used 20 sensors at a 1,000 hertz sampling rate and required 11.4 Mbps JSON. This is quite a relevant test as JSON is a common format to send data to the cloud or across the web.

Without compression or pre-processing mechanisms, an average 7.2 Mbps Internet connection can’t stream data from three or more large machines that require advanced measurement, condition monitoring and traceability of production. A factory must use a connection that is much larger than normal or multiple connections, or it can leverage advanced analytics on the edge.

Edge Devices and Advanced Algorithms

In the past, most programmable logic controllers (PLCs) were capable of handling repetitive tasks in machines, but possessed the computing prowess of a smart toaster. Today’s Industrial PCs (IPCs) feature ample storage and powerhouse processors, such as the Intel® Core™ i7 or Intel® Xeon® offerings, with four or as many as 40 cores. TwinCAT 3 automation software, for example, offers a complete IPC platform that runs alongside Windows, easily supports third-party applications and enables remote access. Most importantly, PC-based control software can provide advanced algorithms to manage data, such as pre-processing, compression, measurement and condition monitoring. This doesn’t require a separate, stand-alone software platform.

Condition monitoring performs many operations locally, such as converting raw accelerometer data into the frequency domain. This can be done on an edge device or within the actual machine controllers’ PLC program. When analyzing vibration, for example, the information is often collected as a 0-10 volt or 4-20 mA signal. This can be changed to a more usable format on the controller through a Fast Fourier Transform (FFT) algorithm. More extensive evaluations of machine vibrations are possible using DIN ISO 10816-3. To monitor bearing life and other specific components, algorithms are readily available to add to a PLC program for calculating the envelope spectrum first and then the power spectrum. Many common machine conditions and predictive maintenance algorithms can be evaluated within the machine control, or on an edge device.

To optimize a Big Data strategy from the ground level upwards, automation software should offer built-in algorithms to process both deterministic and stochastic data. If the data is deterministic, controllers using pre-processing algorithms could send certain values only upon a change, so the recipient should know the mathematic correlation and be able to reconstruct the original signal if desired. For stochastic data, the controller can send statistical information, such as the average value. Although the original signal is unknown, the recipient can still use compressed, statistical information.

It also is possible to implement algorithms on the IPC to monitor process data over a set sequence. This includes writing input data periodically, according to a configured number of learned points, to a file or to a database. After storing standard values, such as torque for a motion operation, algorithms compare cycle values against them. Ensuring the data are within a configured bandwidth creates a type of process window monitoring, which can readjust immediately since the local controller reacts in real-time.

Big Data on Both Edge and Cloud

Running advanced algorithms on a local edge device reduces cloud bandwidth requirements and offers an efficient solution for process optimization guided by Big Data. However, that does not mean a packaging plant should disconnect from the cloud. In the age of IIoT, it is essential to gather and easily access data across an operation, even if many analysis and decision-making tasks can be completed on local hardware first.

To decide what needs to be sent to the cloud and what can be processed or pre-processed locally, make sure to ask a few key questions. First, what are the goals your operation wants to achieve through data acquisition in this instance? Next, which data sets from which machines need to be analyzed in order to achieve these goals? Finally, what types of data insights does the operation need to improve efficiency and profitability?

Local monitoring with edge computing often works most efficiently to improve the operation of individual machines. However, the cloud provides the best platform to compare separate machines, production lines or manufacturing sites against each other. Implementing both allows an operation to maximize its Big Data strategies.

The post Augment Big Data Strategy with Advanced Analytics appeared first on Artificial Intelligence.

Businesses are now collecting and using a huge amount of data. Much of this flows from an increasing range of smart devices, all interconnected as the IoT (Internet of Things). Our computer capacity continues to grow rapidly, but there is still concern over security issues that can compromise even locally generated information. In business, highly sensitive information is stored, and it’s necessary to observe government regulations to protect consumers. At the same time, incidents of data breaches continue to rise. This is why it’s essential to make data protection a priority and establish strict security measures.

Here are five tips that can help you guard data against breaches in both big data deployments and any software accessing the data.

Secure Data Storage

Managing storage is a critical part of any data strategy. Auto-tiering is necessary when you’re looking at petabytes of information. New data is automatically assigned to different storage levels to make managing huge volumes of data simpler. However, this can create other problems due to issues like unverified services or contradictory protocols. Auto-tiering also generates logs of its storage activities which also have to be protected and maintained. SUNDR (secure untrusted data repository) helps with this by monitoring for and detecting unauthorized file operations. These can come from malicious software agents. SUNDR utilizes consistency checks to ensure data is stored securely.

Secure Non-Relational Data

Many organizations handle a large volume of unstructured data such as images. They turn from standard SQL relational databases to NoSQL deployments. These solutions are becoming more common but are still vulnerable to injection attacks where malignant code is inserted. Recommended security measures include hashing or encrypting passwords. You should also use effective end-to-end encryption algorithms such as RSA, AES, and SHA-256, as well as SSL encryption.

Ensure Endpoint Security

Trusted certificates at each endpoint will help to ensure that your data remains secured. Additional measures that your organization should use include regular resource testing and allowing only trusted devices to connect to your network through the use of am MDM (mobile device management) platform.

One challenge lies in ensuring that all data is valid, given the wide scope of devices and data collection technologies. Many input devices and applications are vulnerable to hackers and malware. Intruders can mimic multiple logon IDs or corrupt the system with false data. Your big data solution should be capable of both preventing intrusion and identifying false data.

Prevent Inside Threats

Your company is also exposed to internal security risks, whether from disgruntled or simply careless employees. This is especially challenging in business environments where employees working with the data are not fully educated on proper security practices and behavior, including data scientists and software developers.

It’s important that you provide digital security training to all employees. They should know about password safety, logging off unused computers, granting permissions to other employees and risks of accessing data via public Wi-Fi.

Your company should also have user logs in place to help identify workers who might attempt to steal intellectual property, use stolen logon credentials, or otherwise try to compromise or bypass network security protocols.

Analyze and Monitor

A big data solution that includes tools for both analysis and monitoring in real time can raise alerts the instant network intrusion is detected. But this can result in large amounts of network data. Your goal is to provide an overall picture of what’s happening over sometimes large networks from moment to moment. Your organization may not have the resources to monitor and analyze all the feedback generated, including false alarms as well as real threats.

The solution is that big data analytics itself can be used to improve network protection. Your security logs can be mined for anomalous network connections. This will make it easier for you to identify actual attacks as opposed to false positives going forward.

Final Thoughts

Automated data collection is increasing the exposure of companies to data loss. When considering a big data solution, you can best mitigate the risks through strategies such as employee training and varied encryption techniques. But it’s also crucial to look for solutions where real security data can be analyzed to drive improvements.

The post 5 Tips to Overcome Big Data Security Issues appeared first on Artificial Intelligence.

IT teams can salvage the value of legacy systems while pivoting to a new foundation offering better customer experiences.

CIOs and their teams are routinely maintaining costly, long-running legacy systems that have gathered a large amount of data about purchases, services, and customer behavior over the years. Sure, they’re getting a bit creaky in old age; but the most important problem has existed since they came online — each one was built in a vacuum and effectively siloed away from the others.

IT teams are feeling the strain of maintaining legacy systems at a time when customers are demanding a more tailored, robust, and streamlined experience from their organizations. Today, most customers want insights and capabilities to be accessed quickly and easily, whenever and wherever they need them, while receiving real-time updates and information through multiple channels (for example, web portals, mobile applications, e-mail, SMS). Because various older systems were one-off development projects that didn’t consider future dependencies or integrations, these legacy infrastructures simply aren’t prepared to meet the demands of the 24/7 digital economy without a lot of time and money being invested.

As a result, most organizations with legacy systems are losing revenue that their current offerings can’t drive or, even worse, are watching it get directed towards their competitors. New experiences that provide valuable insights and new capabilities are the best way for an organization to honor customer expectations and remain relevant in most market spaces. But here’s the problem: The data and logic necessary for these new applications is broken up across their legacy platforms, making it very difficult to access the full spectrum of information or the full range of capabilities needed to invent something new and valuable.

Fortunately, IT teams can salvage the value of legacy systems while pivoting to a new foundation. Core back-end scalable architecture supports quick and cost-effective new development without creating unnecessary technical debt. To enable those sought-after next-generation experiences, it takes a lean, thoughtful, and flexible service-oriented architecture that brokers access to both new and legacy information and services. Building this takes extra planning, talent, and effort that many IT teams aren’t ready to commit to, but here are a few key steps to make the process easier and ultimately more successful in the end.

Devise a strategy that allows legacy systems to participate in the new world. While it’s easy to imagine a new greenfield approach that will free your teams from legacy woes, it’s not realistic. Any solution needs to include an integration strategy for legacy systems that keeps them in the mix in the short term while eventually updating or replacing them when the time is right. Try to boil the ocean and you’ll get burned.

Ask the hard questions up front. IT teams must examine their current infrastructure and data assets to identify what needs to be changed to meet their objectives, then define the right technical architecture to meet the needs of the business. While it is tempting to focus on the immediate projects at hand, IT teams must consider the long-term initiatives that will follow in the months and years ahead. Systems, hardware, data, and personnel resources all need to be factored into the plan. To do so, they must ask themselves these tough questions:

• What are the problems you’re trying to solve for your business or your customers, now and in the future?
• What are the desired business outcomes that your technology would power or enable?
• How are you going to measure the success of your initiatives?
• What software, hardware, data, people, and other resources do you have in place now? Do they have the skills you need or should you hire new personnel or bring in outside help?
• What is the technology architecture you will need to facilitate the outcomes you’re after? To the previous point, are you sure that you have the right people in place to answer this correctly?
• How can you bridge the gaps between where you are now and where you need to be?

Once you have a better grasp of where you’re at, examine where you need to be from an overall enterprise architecture standpoint.  Consider a services-based approach that will drive business outcomes over several years. Here are some of the most important elements to consider:

Microservices.  A microservices architecture provides focused and independently deployable application components that fulfill three key objectives: development agility; deployment flexibility; and precise scalability. The highly granular, purpose-built nature of a microservice also facilitates a progressive migration strategy by adding to or replacing legacy components in smaller, more manageable pieces.

Big data strategy. “Big data” is so routine now that it’s better to just call this your “Data strategy.” New technology initiatives will usher in new data dependencies that you may not be accustomed to handling in your organization. As that data is processed and stored, you’ll need a consistent and rapid way to interact with it that will scale as your organization grows and the demands on your data architecture continue to evolve.

Security. Factor security into your plans from the start. You’ll want a straightforward security model that lays across infrastructure – covering data, service, and front-end tiers. Ensure that you have full customization of roles and permissions to account for the various types of users who will be working with your applications now and in the future.

Cloud support. This is also a good time to consider infrastructure flexibility. If you don’t have a cloud strategy, take a hard look at why not. Cloud providers, such as Microsoft Azure or Amazon Web Services, can provide instant geographical distribution, high availability configurations, elastic scaling, and several other useful services beyond the basics, for less money.

Production ready. Design your infrastructure with a focus on centralized monitoring, auditing, and continuous release. Your fancy new application ecosystem is worthless if you can’t manage it. You need to easily diagnose issues as they arise, and ideally have enough monitoring in place to recognize things before they turn into issues. Your system should tie notifications to workflows to prevent troubleshooting downtime and ensure that the right people are receiving the right information at the right time.

DevOps. You need deliver new functionality and fixes early and often in a reliable way, and for that you will need DevOps. Quality analysis, automated tests, packaging, and deployment should be a well-oiled machine with high levels of automation.

Lastly, be certain that your strategy achieves short- and long-term business objectives from the start, rather than doing it piecemeal. You’ll save both time and money by avoiding costly adjustments along with way. You’ll also preserve the sanity of your IT teams and development partners in the process. When you can show immediate ROI on your first initiative, you’ll develop a tail-wind of support from the business that will allow you to realize the long-term roadmap.

The post Set the Right Foundation for Applications with Improved CX appeared first on Artificial Intelligence.