Computer Science Research Opportunities in Sustainability

of 26
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report
Category:

Documents

Published:

Views: 46 | Pages: 26

Extension: PDF | Download: 0

Share
Description
Computer Science Research Opportunities in Sustainability. Randal E. Bryant. http://www.cs.cmu.edu/~bryant. Background. Workshop on the Role of Information Sciences and Engineering in Sustainability Sponsored by NSF, run by Computing Community Consortium February 3-4, 2011
Transcript
Computer ScienceResearch Opportunities in SustainabilityRandal E. Bryanthttp://www.cs.cmu.edu/~bryantBackground
  • Workshop on the Role of Information Sciences and Engineering in Sustainability
  • Sponsored by NSF, run by Computing Community Consortium
  • February 3-4, 2011
  • ~ 60 participants
  • Organizers: Randal Bryant, Doug Fisher, Erwin Gianchandani, Carla Gomes, William Rouse, Prashant Shenoy, Robert Sproull, David Waltz, and Krishna Kant
  • Report
  • Available at: http://cra.org/ccc/seesit_report.php
  • NSF SEES Program
  • Science Engineering and Education for Sustainability
  • Spans entire NSF
  • Budget
  • FY 2012 request: $998 million
  • CISE request: $46 million
  • Sustainability
  • “Development that meets the needs of the present without compromising the ability of future generations to meet their own needs”
  • --Brundtland Commission of UN, 1987
  • Some Dimensions of SustainabilityBuildingsLow Power ITSmart GridEnergyElectricityRenewable energyTransportationElectric VehiclesRide SharingEnvironmentTraffic OptimizationClimateHabitat PreservationAdapting toclimate changeGoogle Data Centers
  • Dalles, Oregon
  • Hydroelectric power @ 2¢ / KW Hr
  • 50 Megawatts
  • Enough to power 60,000 homes
  • Engineered for maximum modularity & power efficiency
  • Container: 1160 servers, 250KW
  • Server: 2 disks, 2 processors
  • IT and Energy
  • Data Center Power
  • The research topic of choice for many computer scientists
  • Interesting problems, lots of progress
  • ~2% of US power consumption
  • Beyond Data Centers
  • How can IT fundamentally improve the processes of electricity generation, transmission, and consumption?
  • Other aspects of sustainability
  • How can computer scientists contribute?
  • With ideas & approaches that our counterparts in electrical engineering, mechanical engineering, & civil engineering would not think of
  • Computational thinking for sustainability
  • Advanced IT’s Role in Sustainability
  • Systems that continuously monitor themselves, and adapt, repair and optimize
  • Systems designed as large networks of loosely coupled agents
  • Serving the needs and characteristics of people
  • Trustworthy modeling and simulation
  • Self Monitoring, Adapting, Repairing, and Tuning Sensors EverywhereAdaptive SystemsTurn HVAC & lights on & offDiagnose faulty appliancesAdjust wind farms to weather patternsRoute around congestionDetect and report environmnetal risks
  • Building occupancy
  • Localized energy usage
  • Energy generation
  • Traffic flow
  • Environmental monitors
  • Sensor-Rich Systems
  • Challenges (and Research Opportunities)
  • Configuring
  • Need systems to self-identify and configure themselves
  • Ensuring Reliability
  • What happens when sensor is faulty?
  • Maintaining Privacy
  • Will collect data about people that should not be released
  • Where people are
  • What they are doing
  • Agent-Based Systems“I need 10KWh by 7am tomorrow”
  • Systems continuously negotiate supply & demand with each other
  • Millions of agents working together
  • UtilityCompanyHomeEnergyManager“OK, will provide 2KW from 12:00 to 5:00 at $0.07/KWh.”Agent-Based Systems
  • Motivation
  • Only way to ensure scalability
  • Enables individuals to make their own choices
  • Vs. centralized control schemes proposed by utilities
  • E.g., utility can provide cost incentives to reduce electricity usage during peak loads, rather than shutting down appliances
  • Challenges
  • Ensuring robust operation
  • Despite unexpected events, software errors, …
  • Minimize vulnerability to malicious attack
  • The Role of People
  • Individuals
  • Motivating them to conserve
  • Typical household pays more for TV cable service than for electricity
  • Most individuals do not want to expend lots of time & effort managing resource usage
  • Systems must be highly automated and easy to use
  • Learn individual preferences and apply them
  • Groups
  • Social networking technology can support conservation
  • Exerting subtle forms of peer pressure
  • Sharing of resources, e.g., carpooling
  • Reliable Modeling & Simulation
  • Modeling & Simulation Critical Tools
  • Predicting effects of decisions
  • Current Methods Error-Prone and Lack Transparency
  • What are effects of simplifying assumptions
  • Reliable Modeling & Simulation (cont)
  • Current Methods Error-Prone and Lack Transparency
  • Modeling decisions embedded in complex software
  • Fail to consider adaptations by people
  • E.g., building more freeways causes people to move further away from work
  • Must gain trust of citizens & policy makers
  • Constructing Models Labor Intensive
  • E.g., deriving building performance model from construction diagrams
  • Information rapidly becomes out of date
  • Opportunities
  • Automatically matching model parameters to sensed data
  • Ongoing Research Projects
  • Understanding Climate Change: A data-driven approach
  • Vipin Kumar, U Minnesota. NSF Expedition
  • Analyze climate-related data from satellites, ground-based sensors
  • Relate to predictions generated by climate simulations
  • Ongoing Research Projects (cont.)
  • Institute for Computational Sustainability
  • Carla Gomes, Cornell. NSF Expedition
  • Using data mining & optimization techniques to optimize resource management
  • Glacier ParkSalmon-SelwayYellowstoneOngoing Projects (cont.)
  • Center for Computational Learning Systems
  • David Waltz, Columbia. Mostly industry funding
  • Applying machine learning / data mining to real-world problems
  • Close working relationship with Consolidated Edison
  • Data mining of maintenance records
  • Placement of EV charging stations
  • Analysis of 2003 blackout
  • Some Characteristics of Successes
  • Use-Inspired Research
  • How CSE can be applied to non-IT problems
  • Close engagement with specific problems
  • Close Collaboration with Domain Experts
  • Academics in other fields
  • Industry
  • Federal agencies
  • Scale of Operations
  • Easier to do as part of large-scale, interdisciplinary center
  • But, there are instances of success at single-PI level
  • Smaller-Scale Examples
  • Formally Verifying Distributed Vehicle Control System
  • Andre Platzer, CMU
  • Based on model of system to control braking, acceleration, and lane changes by autonomous vehicles
  • Smaller Scale Examples (cont.)
  • Low-cost sensors for monitoring home resource utilization
  • Shwetak Patel, U Washington
  • Sensors + Machine learning + HCI
  • Smaller Scale Examples (cont.)
  • Improved Power Control for Electric Vehicles
  • Illah Nourbakhsh, CMU
  • Use machine learning & crowd sourcing to optimize energy management
  • BatteriesMotor / GeneratorSuper CapacitorExploiting Super Capacitor
  • Super Capacitor acts as Power Cache
  • Absorbs short-term fluctuations in charge / discharge
  • Reduces stress on batteries
  • Control Strategy
  • Want to charge when braking for stop sign / discharge when start
  • How can vehicle predict upcoming charge / discharge needs?
  • ControllerBatteriesMotor /GeneratorSuperCapacitorBenefits of Super Capacitor
  • 100 W/h storage ≈ 11ml gasoline
  • Yields 37% savings on battery duty
  • Working with U.S. Industry
  • Highly fragmented
  • E.g., Somerset County, PA windfarm
  • Developed by Atlantic Renewable Energy Corp & Zikha Renewable Energy LLC
  • GE Wind turbines
  • Operated by Florida Power & Light
  • Supplied into grid by FirstEnergy Solutions
  • Who could manage comprehensive overhaul?
  • Low R&D Investment
  • Regulated monopolies have limited incentive to innovate
  • High capital costs & safety concerns limit willingness to make drastic changes
  • Conclusions
  • Research Opportunities Across Many Areas of CSE
  • Cyberphysical systems
  • Systems engineering
  • Machine learning & data mining
  • Optimization
  • Agent-based systems
  • Modeling & simulation
  • Human-computer interaction
  • Seek Participation From Across CSE Community
  • Report
  • http://cra.org/ccc/seesit_report.php
  • Recommended
    View more...
    We Need Your Support
    Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

    Thanks to everyone for your continued support.

    No, Thanks
    SAVE OUR EARTH

    We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

    More details...

    Sign Now!

    We are very appreciated for your Prompt Action!

    x