The Future of Distributed Intelligence with 5G and AI

Subscribe To Download This Insight

3Q 2020 | IN-5885

This ABI Insight aims to explore the potential of 5G and Artificial Intelligence (AI) in enabling distributed intelligence. While there is still a long way to go, the industry is taking the right step. To achieve this reality, the industry needs to appreciate the importance of compute and connectivity at every level and focus on use cases that can take advantage of a wide spectrum of processing locations, from edge AI to centralized processing in the cloud.

Registered users can unlock up to five pieces of premium content each month.

Log in or register to unlock this Insight.


What Constitutes Distributed Intelligence?


Distributed intelligence is an architecture that breaks computation workloads hosted in a centralized system into a combination of one central node and multiple end nodes. The end nodes generally involve multiple machines in many locations and different form factors and processing capabilities, ranging from smart sensors and intelligence devices to gateways and on-premises servers. The idea of distributed intelligence has always attracted many technology solution suppliers, as it provides the following benefits:

  • Quick Response Time: Addressing a task through a combination of one central node and multiple end nodes is very critical in the age of cloud computing. The central node that resides in the cloud is responsible for task allocation, resource scheduling, and optimization. Smart end nodes, on the other hand, can take immediate action without having to exchange information back and forth with the central node, which ensures low latency.
  • Collaborative and Coordinative System: The end nodes can be designed to be aware of each other, so they can work together toward a common goal. The end nodes can also learn from each other and share information to make the entire system more responsive and intelligent.
  • High Redundancy: An architecture with multiple subsystems eliminates Single Point of Failure (SPOF). Sharing workloads across multiple endpoints reduces the processing demands and complexity of each individual machine as compared to a monolithic system.

While this architecture has greatly benefited the design and implementation of various systems, such as cloud computing clusters, warehouse robots, and smart home systems, these systems are often limited by geographical factors, connectivity options, and the processing capabilities in end nodes. The emergence of 5G and Artificial Intelligence (AI) is set to change these. The ubiquitous presence of both technologies in the connectivity and processing layer will usher in a distributed system that can serve a wide range of use cases with high flexibility and scalability.

The Impact of 5G and AI


The year 2019 was the dawn of 5G era. Equipped with high throughput, low latency, and massive Internet of Things (IoT) connectivity, 5G is primed to serve non-consumer markets. As more and more telcos and private network service providers launch 5G networks, they are looking to target key verticals such as industrial and manufacturing, transportation and logistics, healthcare, and retail. This means there will be a wide range of business-critical and mission-critical devices getting connected to the 5G network. Currently, these IoT devices have varying levels of computing capabilities, but the democratization of AI is likely to change that.

In recent years, the AI industry has been actively expanding the coverage of AI beyond cloud-based processing. Compared to cloud-based AI, which focuses on recommender systems, conversational AI, and time series financial forecasting, edge AI focuses on introducing AI to smart devices. Through either connecting these devices to gateways and on-premises servers or embedding AI chipsets into smart devices and smart sensors, 5G-enabled edge AI devices will have the flexibility to centralize all their workloads in the cloud, or perform time-, latency- and security-sensitive workloads at the edge. This removes data privacy, safety, and security concerns while allowing the overall system to update and optimize itself.

Given their impact, ABI Research forecasts that the total output generated by AI and 5G combinations to the global economy will reach US$17.9 trillion, or 9.7% of the global Gross Domestic Product (GDP), by 2035 in our 5G and AI: The Foundations for the Next Societal Leap Whitepaper. Key application areas that benefit from both 5G and AI include flexible assembly lines at wireless factories, last-mile delivery robots and drones, remote surgery theatres in hospitals, and Augmented Reality (AR) glasses in field service. These devices are mobile, untethered, and battery-powered and engage in business-critical operations. 5G and AI can ensure these devices have the capabilities to process information on the spot or exchange key information with the host system in the cloud, regardless of their physical location.

Actionable Insights for Key Stakeholders


To truly enable compute and connectivity at every single layer of a distributed intelligence system, technology vendors must focus on the following:

  • Telcos should consider an AI-as-a-Service (AIaaS) approach, whereby they could use the technology in line with 5G to differentiate their value propositions to end customers. Telcos must invest in new telco infrastructure and service delivery models, such as 5G standalone, multi-access edge computing, and telco AI, early and boldly to enable network slicing and Service-Based Architecture (SBA) that will deliver AIaaS to their end users. They will also need to conduct Proof of Concept (PoC) and invest in solutions that focus on enterprise.
  • Chipset vendors must pay attention to new emerging hardware trends, such as Tiny Machine Learning (TinyML) and Silicon-as-a-Service (SiaaS). End users are constantly pushing the boundaries of edge AI, with the aim to bring AI into as many sensors and sub-components as possible. Since partnership and ecosystems become very important in enabling wide-scale adoptions, chipset vendors must work closely with the device manufacturers, Systems Integrators (SIs), and software developers in open source communities to provide support and incentive while facilitating the rollout of innovative products that will take advantages of both 5G and AI.
  • Regulators, industrial associations, and standardization bodies such as the 3rd Generation Partnership Project (3GPP), the European Telecommunications Standards Institute (ETSI), and TM Forum should not wait on this trend for too long. They must engage with key stakeholders across the value chain and access current regulatory gaps in the deployment of 5G and AI. Setting up the right ethical framework, for example, is extremely critical to ensure these technologies are transparent, explainable, and ethical, so that they can be easily accepted by mass consumers.

5G and AI are the most innovative technologies in the market today. Their combination will allow networks and infrastructure to be more efficient, which will have a significant impact on business productivity while optimizing overall network resources. 5G will initially accelerate the development of AI applications before ushering in distributed intelligence throughout the network. Finally, 5G and AI will create completely new service paradigms.



Companies Mentioned