Professor Ian Leslie

PhD
FREng
Fellow
University Professor of Computer Science
President 2014-2019
Subject
Computer Science
Email
iml1@cl.cam.ac.uk

My research interests have been in computer networks and operating systems - particularly with respect to performance guarantees. More recently I have focused on the use of information systems to reduce energy demand. This has led to two different campaigns.

The first campaign is about making energy information more readily available. This involves efforts to drive down the cost of fine-grain monitoring, making the raw information gathered publicly available, and developing open-source tools that can interpret such information. Some of this can be seen in the Gates Building energy visualistion<http://www.cl.cam.ac.uk/meters/vis> (click on the nodes on the tree at the left to get started). The low cost monitors described here<http://www.cl.cam.ac.uk/research/srg/netos/c-aware/hardwaremeters.html> are now being piloted in the Department of Pathology at Cambridge. The same visualisation tool can be used to look at this information - currently rather nascent - at the Pathology energy page<http://www.path.cam.ac.uk/energy/vis>.

The visualisation tool(s) continue to be developed, but the more important point is that the information used by the tools is publicly available; anyone may develop tools to process the data in better ways.

The other campaign comes from a dual realisation that (i) we lack information on occupant behaviour that we might use to contextualise energy use, and (ii) that the way we distribute power for lighting is badly out-of-date. Tbese seemingly disconnected notions jointly give rise to a DC powered lighting and sensor network The key ideas are a 24VDC network, battery backed up, easy to integrate renewables (particularly solar), convenient power for processors and sensors (and LEDs), much higher reliability and efficiency. This is open platform - not just about lighting - and the long component lifetime means that putting a processor alongside every light is sensible, As you can see from the Gates Building energy visualisation, 20% of our electricity use is on lights. We have a single corridor on a DC network with LED lighting that reduces the energy by about a factor of ten. The computer science starts when we begin to gather information from the sensors co-located with the lights.