Telco edge compute use case: autonomous vehicles

Featured Telco Edge Compute Use Case

Introducing a framework for characterising telco edge compute use cases

STL Partners has been working across Consulting and Research to define the opportunity for telco edge computing and associated use cases. We see ideal candidate use cases for edge computing as those demonstrating a balance of requirements characteristic of both local (on-device / on-premises) computing (low latency, reduced backhaul and data localisation) and cloud computing (mobility, scalability and light device). This framework is best explored using an example edge use case, as described below.

Figure 1: The local vs. cloud balance

Source: STL Partners

Use case: Autonomous vehicle – intersection movement assist (IMA)

In this use case, telco edge computing can be used to support one of the types of safety applications in autonomous driving. In the scenario, vehicles approaching from all sides of an intersection with limited visibility of each other communicate between themselves to warn drivers of potential collisions. The calculation of a potential collision can be offloaded to a telco edge compute, which may be located at a base station, aggregation point or, in a more mature market, on dedicated telco-run Roadside Units (RSUs).  Facilities sharing by operators would further help to minimise potential latency. Below we have highlighted how this use case maps to our framework for requirements characteristics. This is a clear example where an application would benefit from a solution that incorporates both local compute-like requirements and the flexibility of the cloud.

Figure 2: Autonomous Vehicle: Intersection movement assist (IMA)[1]

Source: STL Partners

Why edge?

The table below uses our framework for characterising telco edge compute use cases to demonstrate how edge compute can enable this type of application and which benefits are most relevant.

Source: STL Partners


[1]
Use case discussed in http://5gaa.org/wp-content/uploads/2017/12/5GAA_T-170219-whitepaper-EdgeComputing_5GAA.pdf