It’s Strategy Time for Edge Computing

Ask the same companies their strategies for edge computing, and the answers will generally be more ambiguous and vary depending on organizational role of the respondent. However, change is afoot, and consistent company strategies for edge computing are emerging.

Communication networks have always had edge computing, but this computing has traditionally been housed in vertically integrated platforms, which are managed independently. As networks become virtualized and particularly as CSPs implement cloud native solutions based on microservices, edge computing architectures and operations will grow in strategic importance. This will be bolstered by enterprise IT initiatives to complement cloud with edge computing and emerging consumer and enterprise services that depend on edge computing capabilities. Notable examples include, augmented, virtual and mixed reality, real-time industrial operations, autonomous and assisted driving, smart-city infotainment and local data management for security and privacy.

Companies are Jockeying for Competitive Advantage

Edge computing is not just the domain of CSPs. Many other companies are making strategic investments with the aim of cornering market opportunities. With these investments, companies naturally play to their strengths, such as service incumbency, real-estate access and ownership and network connectivity. For example, public cloud service providers are developing integrated edge computing capabilities to complement their cloud services. Telecom tower companies are configuring their sites to support edge compute infrastructure, and in some cases are investing in dark fiber assets for network connectivity. Many CSPs are implementing multi-access-edge computing (MEC) to capitalize on their network incumbency and some are pursing initiatives like CORD (Central Offices Re-Architected) to leverage their extensive central office real estate. CSPs have several factors in their favor, including:

• Consumer and enterprise customers that are likely to have future edge computing service requirements;
• Communication networks that are inherently distributed, with emerging edge computing demands for infrastructure virtualization;
• Customers requiring network connectivity services for edge computing applications.

Although market competition for edge computing is inevitable, there are opportunities for most players who are vying for a piece of the action. However, market scale is crucial, particularly for players who dedicate large capital investments for real estate and infrastructure. Furthermore, given the complexities of edge computing, it is inconceivable that an individual player can go alone. The winning players will be those that scale with effective partnership and coopetition strategies.

Agility is King

As CSPs and other companies develop their strategies for edge computing, they must anticipate changes in business cases and service demands as markets develop. These changes will have a direct impact on the edge computing architectures and operational models needed. For example, many initial edge compute implementations will have low utilization rates when implemented for isolated services and will only be viable with infrastructure sharing arrangements. Service demands will increase, and the optimal placement of infrastructure will change as edge computing becomes more readily available and new services are developed. In addition, the success of public cloud computing has demonstrated the tremendous benefits of self-service capabilities. These capabilities are essential for edge computing and can only be achieved with highly automated systems to abstract operational complexities and enable agile services.

Now is the Time to Take the Plunge

Although there are still many uncertainties for edge computing, technical and commercial strategies are coming thick and fast. And they must. As market uncertainties are resolved the opportunities for new entrants will shrink. In the face of these uncertainties, company strategies must support agile architectures and operational models and cater for diverse commercial relationships from the outset. Concepts from other industries, such as fail-fast and minimum viable products might be helpful and extensive automation is essential.