Edge Intelligence: The Central Cloud is Dead – Long Live the Edge Cloud!
software-defined networking principles.
This allows customers to reduce costs
and increase efficiency in a manner
similar to the benefits provided by cloud
computing, as customers do not have to
provision and maintain data centers at
the edge.
The best user interface is no user
interface. Traditionally, user interfaces
enable users to input data, analyze data
and execute decisions. IoT makes
manual data input largely obsolete, as
data is collected automatically from
sensors. Machine learning and artificial
intelligence take over the data analysis
and decision-making. Human interaction
and interference are largely reduced.
The following recommendations are
addressed to the various industries that will
be impacted by the development and
implementation of edge intelligence
applications:
a service, platform as a service, and
infrastructure as a service. These
services are typically located in the cloud,
i.e. in centralized data centers. With the
advent of edge computing, we will see an
extension of these service models to the
edge and combinations of traditional
license and service models.
Utilize 5G standards to facilitate edge
computing and edge intelligence
solutions. 5G networks have the
potential to provide benefits of secure
and scalable network connectivity.
Agree on a common approach to
orchestration
and
life-cycle
management and to machine learning
(tools, model implementation) to avoid
market
fragmentation.
Their
commoditization will drive down cost.
Prepare for disruption of business and
commercial models. During the last
decade, we have experienced a change
from the traditional software license
model to the services model: software as
- 30 -
September 2017