Deep Learning Practices and Challenges in IoT ~ Realization of Intelligent Networks~
Abstract
Preferred Networks Inc. (PFN) is a Tokyo-based startup, focusing on
applying latest deep learning research to emerging problems in the Internet
of Things (IoT).
Intelligent industrial robots (*1), and autonomous robot
car control demonstration in CES2016 (*2) show some of the advanced
technologies developed in collaboration with world-leading partners such as
FANUC and Toyota.
"How to make networks intelligents?" is a key challenge
that PFN has been focusing on since its inception.
In our session, we introduce some deep learning applications in IoT,
focusing on "intelligent networks" aspect.
By focusing especially on the
communication network implementation, we will discuss how we can introduce
"Intelligence" to the network as whole, rather than enhancing the network
devices only.
Based on the above applications, we will discuss some of the problems and
expectations in IoT from the network technology perspective. Below are some
examples of discussion topics:
The limitations of current network technology. Expectations for next
generation technologies such as 5G and LPWA.
Standardization of various IoT connectivity specifications, and their
applications.
We hope that the session will trigger some meaningful discussions around
the future of IoT applications.
