Dawn of the Edge

A case can be made that humanity is on the verge of a new era which will be known as the Data Age. It is characterised by networked electronic devices that incessantly produce an enormous stream of data, autonomous vehicles, humanoid robots and smart personal assistants. It’s also the age of smart home devices and surveillance cameras, and fully-automated manufacturing. In a nutshell, it’s the era when the Internet of Things comes into its own.

Forecasts predict enormous growth. By 2023, the latest Cisco Annual Internet Report anticipates that there will be some 2 billion devices or connections linked to IP networks in Central and Eastern Europe alone—an increase of two thirds compared to 2018. In the same period, the number of networked devices around the world will have increased to close to 30 billion. Half of these will be machine-to-machine (M2M) connections meaning devices, vehicles and facilities that share data between themselves or with central systems. IT is rapidly becoming omnipresent and the IDC market analysts expect that in 2025, the average person will interact nearly 4,800 times with connected devices every day, which roughly equates to once every 18 seconds.

When you harmonise networks, IT departments can manage them through a single pane of glass, leveraging the same functionality and security mechanisms. This makes it easier to control networks, introduce IoT-capable routers and switches, and substantially improve transparency.

Sven Glüsing, Head of Network & Security Solutions, Bechtle Darmstadt

THE VALUE OF DATA.

All of these devices produce, receive and process data. In the next five years, it’s estimated that data generated around the globe will triple to 175 zettabytes. That’s equivalent to 175 billion 1 TB hard drives. However, all of this data and related information have a best before date, as information is only valuable if it reaches the recipient in time to trigger a predetermined reaction.

According to an Industrial Internet Consortium model based on various application scenarios, the value of information about a machine failure in a production plant, for example, drops to zero percent after just one minute, after which the machine may only have scrap value. If you employ predictive maintenance, this can extend the time you can afford to respond and information starts to lose its value after a day, and doesn’t drop to zero percent until after a month. This gives service staff more time to prevent disruptions.

In 2025, the average person will interact nearly 4,800 times with connected devices.

“Data transfer speeds are crucial for businesses who are looking to network and optimise their production facilities”, says Raphael Mintgen, head of Competence Centre: IoT Solutions at Bechtle Cologne/Bonn. “Besides latency and available bandwidth, the connection to where the data are being processed also plays a decisive role.” But in most industrial IoT scenarios, data transfers to the cloud or the company’s in-house data centre are either too slow or there are bottlenecks because of the sheer volume of data continuously being generated by sensors and cameras. This why data processing is shifting closer to where data are being generated—the edge of the network. “In a perfect world, data packages are collated at the machine and reduced down to the most relevant information before being forwarded to parent analytics tools and IoT platforms”, says Raphael Mintgen.

FROM THE EDGE TO THE CORE.

A new computing model is needed to handle the volumes, variety and speed requirements of IoT data. Information is not required equally fast for every scenario, and so data and related information are shared between the sender and recipient using a special network topology according to the specific application. A fundamental, if not the most important function is performed by the middle layer, which is typically referred to as the edge. This layer is sandwiched between the central computer instance—the cloud or the company data centre—and the local layer made of physical hardware such as PCs, smartphones, controllers, sensors, actuators and autonomous vehicles.

Compute, storage and network functionalities that collect, analyses and transfer data—often in real time—to the optimal processing location are consolidated in the edge layer. The portfolio of hardware used ranges from IoT gateways and routers to high-performance rack-mounted servers and micro-data centres in mobile network, factory and vehicle base stations.

Edge computing projects are extremely multifaceted and it takes a good mix of software expertise, detailed knowledge of the IT infrastructure and electronics—depending on the scenario.

Maxim Kushnir, Head of Technology Concepts & Networking, Bechtle Stuttgart

BABY STEPS.

According to a 2019 IDC study, over 40% of the German industrial companies surveyed have piloted or implemented initial IoT projects, while around 50% are still in the planning and evaluation phase. 24% already leverage edge computing and 60% are evaluating its benefits in pilot projects or at least intend to do so.

Bechtle expert, Sven Glüsing, views a largely unified network infrastructure across traditional IT and production-related operational technology (OT) as being crucial for the successful implementation of edge scenarios. “When you harmonise networks, IT departments can manage them through a single pane of glass, leveraging the same functionality and security mechanisms. This makes it easier to control networks, introduce IoT-capable routers and switches, and substantially improve transparency”, says the head of security and network solutions at Bechtle Darmstadt. Leveraging virtualisation and segmentation technologies also makes it possible to keep IT and OT data separate. To protect these interconnected instances from manipulation, data theft and sabotage, tools such as micro-segmentation, access controls and firewalls are all good options especially in the face of a growing number of malware attacks and increasingly sophisticated attempts by cybercriminals to overcome safeguards. “If, for example, a control program is replaced rather than erased, it’s not beyond the realms of possibility that the attack will go unnoticed for a while and the damage caused, the potential disruption to entire supply chains and the high cost associated with ramping up production again can be exponential.”

By 2023, there will be nearly 30 billion devices around the world connected with IP networks.

BABY STEPS.

According to a 2019 IDC study, over 40% of the German industrial companies surveyed have piloted or implemented initial IoT projects, while around 50% are still in the planning and evaluation phase. 24% already leverage edge computing and 60% are evaluating its benefits in pilot projects or at least intend to do so.

Bechtle expert, Sven Glüsing, views a largely unified network infrastructure across traditional IT and production-related operational technology (OT) as being crucial for the successful implementation of edge scenarios. “When you harmonise networks, IT departments can manage them through a single pane of glass, leveraging the same functionality and security mechanisms. This makes it easier to control networks, introduce IoT-capable routers and switches, and substantially improve transparency”, says the head of security and network solutions at Bechtle Darmstadt. Leveraging virtualisation and segmentation technologies also makes it possible to keep IT and OT data separate. To protect these interconnected instances from manipulation, data theft and sabotage, tools such as micro-segmentation, access controls and firewalls are all good options especially in the face of a growing number of malware attacks and increasingly sophisticated attempts by cybercriminals to overcome safeguards. “If, for example, a control program is replaced rather than erased, it’s not beyond the realms of possibility that the attack will go unnoticed for a while and the damage caused, the potential disruption to entire supply chains and the high cost associated with ramping up production again can be exponential.”

No matter if I want to ensure the integrity of a travelling work of art based on ambient temperature and humidity, or that a shipping container remains sealed while at sea, with the right edge devices and a stable connection, a whole range of applications can become a reality.

Raphael Mintgen, Head of Competence Centre IoT Solutions, Bechtle Cologne/Bonn

MUCH MORE THAN SMART MANUFACTURING.

It’s not only industrial process that can be redefined from the edge in, but everywhere current status data can be converted into valuable information. “In a smart city, this could be the ability to make optimum use of parking spaces, in an amusement park to monitor rollercoasters using sensor technology, or to enhance granular tracking of goods in transit. No matter if I want to ensure the integrity of a travelling work of art based on ambient temperature and humidity, or that a shipping container remains sealed while at sea, with the right edge devices and a stable connection, a whole range of applications can become a reality”, says IoT specialist, Raphael Mintgen. Regardless of the industry, more than just the optimisation of existing processes, enriching products with additional data-based services and new as-a-service business models for traditional goods are growing in importance. From subscription laser machines to air filtration systems that process their own status data, digitalisation is also changing the logic of marketing.

A proof of concept is a good way of evaluating an IoT process as both IT staff and the departments that consume the technology work together with experienced partners to align their various requirements and check their feasibility. “Edge computing projects are extremely multifaceted and it takes a good mix of software expertise, detailed knowledge of the IT infrastructure and electronics—depending on the scenario”, says Maxim Kushnir, head of technology concepts and networking at Bechtle Stuttgart. He adds: “Successful testing should always be followed up with a well thought-through and stable operating model.” One thing’s for sure, the edge is coming, but that doesn’t change the fundamental principles of a good IT solution.

Sources and additional reading material:

Cisco Annual Internet Report (2018–2023), 2020
IDC: Deutsche Industrieunternehmen optimieren mit Industrial IoT, Innovation spielt untergeordnete Rolle, 2019
IDC: Data Age 2025. The Digitization of the World – From Edge to Core, 2018
Industrial Internet Consortium: Introduction to Edge Computing in IIoT, 2018