Red Hat’s 2019 IT industry forecasts – Edge computing, blockchain, and artificial intelligence

Red Hat's 2019 IT industry forecasts - Edge computing, blockchain, and artificial intelligence

Blockchain, Edge computing, Internet of Things, artificial intelligence, and machine learning are technologies of the future that are becoming an inseparable part of today’s business world. Experts from Red Hat shed light on them.

1. Blockchain
Ian Hood, chief technologist – Global Service Provider, Red Hat

In 2019, blockchain technology will continue to develop innovative applications across all industries to distribute and secure almost every type of transaction. An increasing part of our professional and private lives will be conducted via public networks and cloud computing. This includes, for example, financial applications. At the same time, users expect that personal data, medical, or legal records will be stored securely.
Consequently, the assumption has emerged that no application trusts any network, as a determined attacker can penetrate any part of the infrastructure. Building security into the development of applications and software from the very beginning will not only improve delivery performance but also impact the quality of security. What we expect this year is an evolution of end-to-end encryption mechanisms used to secure delivery via networks (VPN, SSL, encryption keys). These will become an integral part of application development, with blockchain technology as a key element of these cloud-native applications dedicated to healthcare, telecommunications, and many other industries.

2. Internet of Things
Tom Nadeau, Technical Director, NFV, Red Hat

The importance of the Internet of Things will grow even further over the coming months. We are moving toward expanding network connectivity and access to an ever-increasing number of devices or systems, such as cars, sensors, and other specialized equipment designed for industrial, home, or mobile applications. Below is a breakdown of developing technologies that may dominate the market in the coming months.

a) Business models
Most enterprises are continuing to deploy their chosen model for managing IoT-connected devices, whether they are purchased and managed directly by the company or owned by employees. Some organizations are also considering outsourcing device management, not only by transferring these responsibilities to telecommunications service providers but also by delegating tasks to OTT (Over The Top) services.
Service providers currently provide IoT service management, e.g., remote home alarm systems like Comcast. Others are working with companies to relieve them of the burden of managing these devices. Meanwhile, OTT providers like Google are expanding their offerings by using distributed systems for cloud management, such as Nest.
Still other entities will deal with vehicle electrification and management. Even large agencies, such as the US Federal Aviation Administration (FAA), are considering managing aircraft using similar tools. Managing, or at least monitoring, drones in a similar way is also being considered.

b) Device management
In this area, one of the main challenges remains the lifecycle management and health monitoring of devices. It is worth remembering that those connected to the Internet of Things do not always have access to a reliable network with sufficient bandwidth. This entails the need to adapt these systems to a constantly changing environment. For example, firmware updates should be carried out at the right time to avoid a situation where a device operating at the network edge loses connectivity before the update is finished.
IoT device lifecycle management (including software or firmware) is the subject of many open source projects. These communities focus on a distributed Edge computing architecture that uses projects like OpenStack or Kubernetes. Within these projects, “remote” clusters serve to increase the reach and scale of device management software, whether it is a home alarm system or industrial automation using robots. Newer projects, such as Akraino Edge Stack, are an attempt to address device management issues by creating software stacks from the components needed to run Edge computing solutions.
There are many interesting initiatives that present non-standard approaches to managing the Internet of Things. For example, many projects related to vRAN (virtual Radio Access Network) networks are being created, such as Open RAN (ORAN). Those involved are working closely with network operators to ensure that the proposed solutions are as compliant as possible with existing infrastructure needs.
The effort described above is focused on supporting existing Edge technology use cases, but it also relates to supporting, maintaining, and providing relevant information to IoT devices regarding needs that may arise in the future related to tethering and off-network control. This will also concern implications related to the mass scale of IoT application.

3. Artificial Intelligence (AI) / Machine Learning (ML)
Daniel Riek, senior director, AI, Office of the CTO, Red Hat

Artificial Intelligence (AI) workloads will become a primary driver of IT strategy. The use of AI technology reflects the development and transformation of the IT industry. Contractors are increasingly focusing on intelligent applications that could enable them to improve their business. This applies to any way software works – not just in the scope of traditional business operations, but also in research, production processes, and – increasingly – with regard to the products themselves. The increased scale of automation achievable using artificial intelligence will quickly become critical to building company competitiveness, making AI a technology that defines operational strategy.
Kubernetes (as well as Amazon S3 and open source Apache Kafka) will be recognized as the default platform for scaling artificial intelligence workloads. In terms of platforms, the development of intelligent applications is driving a shift from the traditional Big Data model, as a standalone system, to a convergent, data-centric application platform. Linux and Kubernetes form the core of technology in terms of these changes, along with S3 as an interface for data at rest, where data in motion is handled by Apache Kafka.
AIOps will revolutionize the operations of IT departments, software development areas, and management known as SRE (Site Reliability Engineering). IT processes themselves are a valuable field for the application of AI. The excellence of these processes is of critical importance, especially for achieving success in the modern, software-defined world of computing resources. AI-based automation and optimization based on learning from aggregated performance data will give the processes a new dimension while democratizing cloud operations excellence to a degree that has hitherto been achievable only for the largest native cloud solution operators.

4. Edge computing
Mike Hansen, sales development lead, telco, Red Hat

Information from the Internet of Things and other data collected via wireless networks is currently most often transmitted to a private or public cloud. There, it is processed using analytics, artificial intelligence, or machine learning. The 5G network will allow for the transmission of massive amounts of data to distributed wireless networks, making further transmission of this data to a centralized cloud impractical. Data will be managed and processed locally, at the edge of the CSP (Communicating Sequential Processes) network.
This trend will create an opportunity (and an economic necessity) for wireless service providers to monetize 5G networks by offering data management and processing services, such as analytics, artificial intelligence, and machine learning. In 2019, we will see the formation of a global consortium of wireless service providers. This consortium will work together to develop a common platform and methodology for running “intelligent” applications that incorporate data analytics, AI capabilities, and machine learning at the edge of these service providers’ networks.
We currently talk about hybrid clouds, connecting private and public infrastructure via a WAN, which is often maintained by a telecommunications provider. In 2019, a new paradigm known as Hybrid Edge Computing will also emerge. The Hybrid Edge Computing architecture will represent the third type of cloud in the hybrid computing model: it will be infrastructure located at the edge of the telecommunications provider’s network. This model will be necessary to support “intelligent” applications that use data collected within 5G networks.