Welcome to the IoT Slam® Live 2017, Internet of Things Conference, June 21 - 22 2017, R.T.P - North Carolina, USA

Dynamic Machine Learning and Prescriptive Analytics Have Already Saved 1 Billion kWh in Data Centers

01 Dec 2016
16:05 - 16:35
Radiance Room

Dynamic Machine Learning and Prescriptive Analytics Have Already Saved 1 Billion kWh in Data Centers

Session Abstract:

PrintOver 400 data centers have moved from design-based cooling management to an adaptive AI approach that has resulted in a significant reduction in risk, a 40% reduction in energy spend and CO2 emissions, and a 10-15% increase in capacity.  This session will highlight the architecture and algorithms that delivered these results, and how they can be applied to optimizing other kinds of facilities.

Mission critical data centers are at the core of the knowledge economy.  Yet, more often than you’d think, high temperatures in data centers cost companies millions of dollars in application downtime and lost revenue.  In an effort to mitigate this risk, facility operators invest in and operate cooling equipment using traditional ‘best practices’ that consume almost $8 billion in energy every year.  Even though the traditional approaches are meant to be conservative, they are very risky and inefficient.

This session will show how sensors, machine learning and prescriptive analytics bring data center management into the modern age.  What are the challenges associated with applying machine learning to a system that performs feedback control and changes dynamically over time, and where humans are both in the loop and a significant source of disturbance?  How can machine learning and analytics overcome the resistance that data center operators feel before they will let go of their imagined steering wheel?  How can machine learning best use the billions of collected data points to meet temperature requirements in the most efficient way that is also safe in the most unusual circumstances?  How can analytics prescribe for management how to best invest their limited capital to minimize risk, reduce energy spend and maximize capacity?  All of this will be discussed with actual case study results.


Cliff Federspiel
President & CTO at Vigilent

Dr. Cliff Federspiel is a leader and visionary in the field of energy management, having authored more than 50 papers. He is the recipient of the Ralph G. Nevins Physiology and Human Environment Award from ASHRAE, and holds numerous patents in the field.

Dr. Federspiel began his career in R&D at Johnson Controls after receiving his BSME from California Polytechnic State University San Luis Obispo, and his SMME and PhD from the Massachusetts Institute of Technology. In 1998 he joined the University of California, Berkeley, where he was a Specialist at The Center for Environmental Design Research and the Electronics Research Laboratory.

In 2004 Dr. Federspiel launched Vigilent, building on his pioneering research in dynamic cooling technology and wireless networking, to address the emerging need of treating cooling and energy consumption as a managed resource. Today, he is a leading voice and a frequent speaker in the field of sustainable energy solutions and green technology.

Dr. Federspiel interview at Bloomberg New Energy Pioneer conference in 2016:

Session Tags:

End-User, Government, Enterprise, OEM

Machine Learning, Prescriptive Analytics, Predictive Analytics, Artificial Intelligence, Wireless Mesh Network, Data Center, Telecom, Infrastructure, Industrial IoT

CxO, VP / Director, Middle Management, Technical, Business Line Management, Operations




Join our IoT Community at https://www.linkedin.com/groups/4662022/profile 

IoT Grand Slam Virtual Internet of Things Conference

IoT Slam Internet of Things Conference, Bryson Koehler