Edge Intelligence in the IoT
The proliferation of AI in all aspects of life and work is very visible. Traditional ways of analyzing data and arriving at decisions are being replaced by machine learning (ML). ML has two phases - one, a training phase where a lot of data is used to converge on algorithms to analyze them, and then an inference phase where new data is quickly processed through the algorithm, or 'network', created. In IoT systems, the accent
so far has been on sensing and sending data to cloud servers for processing. The IoT nodes, therefore, were created to efficiently collect and transmit data - the nodes did not have the capability to run either phase of ML algorithms internally. There is a case for carrying out analysis and a first level of decision making at the IoT node or device itself. This 'edge intelligence' concept is spreading fast into all kinds of systems - for
example, face recognition, medical diagnostics, industrial control systems, etc.
This talk provides a background to IoT with Edge Intelligence. It then describes how algorithms can be pushed towards the edge, and how hardware structures can be created to carry out energy efficient computation of ML algorithms or deep neural networks. The talk would cover both hardware and software implementations, and would include performance and power optimization considerations.
Date and Time
Location
Hosts
Registration
-
Add Event to Calendar
- BITS Pilani Hyderabad Campus
- Hyderabad, Andhra Pradesh
- India
Speakers
N Venkatesh of Redpine Signals
Edge Intelligence in IoT
Biography:
N.Venkatesh is Vice President of Advanced Technologies at Redpine Signals, and has 30 years of engineering and management experience in wireless system design, IoT solutions, semiconductor design, and avionics. His current areas of interest are on building semiconductor and system solutions in the field of the Internet of Things with Edge Intelligence.
He is also the current Chairman of IEEE Hyderabad Section, and a Board Member of TiE Hyderabad.
Mr. Venkatesh holds a Masters Degree in Electrical Engineering from the Indian Institute of Technology, Madras, India. He holds 22 US patents, has contributed to IEEE standards development and has written numerous articles in technical journals. He is a Fellow of the Indian National Academy of Engineering and a recipient of the VASVIK Award for Industrial Research in 2011.
Agenda
2:30 PM to 4 PM: Technical talk