Adrian Albert

Steptacular: An incentive mechanism for promoting wellness

This paper describes Steptacular, an online interactive incentive system for encouraging people to walk more. A trial offering Steptacular to the employees of Accenture-USA was conducted over a 6 month period. Over 5,000 employees registered for the program and close to 3,000 participants wore USB-enabled pedometers; from time to time they plugged their pedometer into a computer to upload hourly step counts to a website; and the website had a range of features to encourage more walking. These features included monetary rewards which were randomly redeemable through a simple game, and a social component. We describe the system and present preliminary findings about the effectiveness of each of these features in encouraging physical activity.

Segmenting consumers using smart meter data

Existing electricity market segmentation analysis techniques only make use of limited consumption statistics (usually averages and variances). In this paper we use power demand distributions (PDDs) obtained from fine-grain smart meter data to perform market segmentation based on distributional clustering. We apply this approach to mining 8 months of readings from about 1000 US Google employees.

Finding the Right Consumers for Thermal Demand-Response: An Experimental Evaluation

For demand-side management programs concerned with heating, ventilation, and air conditioning (HVAC) energy consumption, smart meter data collected at the whole-premise level has recently been used to decompose usage into its HVAC and non-thermal components, which are typically not separately monitored. In this paper, we study the extent to which program design and decisions based on models using whole-home energy consumption differ from decisions made with full knowledge of appliance-level end-use patterns. We develop a model assessment methodology for the case when model results are used to rank consumers by their potential for thermal demand-response. For this, we compare rankings of consumers in two scenarios—first, when only the aggregate outcome of the top consumers matters, then when the relative ordering of the consumers is important. We illustrate our methodology using two individual consumption models that extract thermal (temperature-sensitive) and occupant-driven components from single-point source smart meter data. Moreover, we discuss how a demand response program that selects the consumers with the most potential for energy reduction based on model results may achieve similar results as in the ideal case when separately monitored HVAC data is used.

Cost-of-Service Segmentation of Energy Consumers

Uncertainty in consumption is a key challenge at energy utility companies, which are faced with balancing highly stochastic demand with increasingly volatile supply characterized by significant penetration rates of intermittent renewable sources. This paper proposes a methodology to quantify uncertainty in consumption that highlights the dependence of the cost-of-service with volatility in demand. We use a large and rich dataset of consumption time series to provide evidence that there is a substantial degree of high-level structure in the statistics of consumption across users which may be partially explained by certain characteristics of the users. To uncover this structure, we propose a new technique for extracting typical statistical signatures of consumption-energy demand distributions (EDDs)-that is based on clustering distributions using a fast, approximated algorithm. We next study the factors influencing the choice of consumption signature and identify certain types of appliances and behaviors related to appliance operation that are most predictive. Finally, we comment on how structure in consumption statistics may be used to target residential energy efficiency programs to achieve greatest impact in curtailing cost of service.

Cost-of-service segmentation of energy consumers

Uncertainty in consumption is a key challenge at energy utility companies, which are faced with balancing highly stochastic demand with increasingly volatile supply characterized by significant penetration rates of intermittent renewable sources. This paper proposes a methodology to quantify uncertainty in consumption that highlights the dependence of the cost-of-service with volatility in demand. We use a large and rich dataset of consumption time series to provide evidence that there is a substantial degree of high-level structure in the statistics of consumption across users which may be partially explained by certain characteristics of the users. To uncover this structure, we propose a new technique for extracting typical statistical signatures of consumption-energy demand distributions (EDDs)-that is based on clustering distributions using a fast, approximated algorithm. We next study the factors influencing the choice of consumption signature and identify certain types of appliances and behaviors related to appliance operation that are most predictive. Finally, we comment on how structure in consumption statistics may be used to target residential energy efficiency programs to achieve greatest impact in curtailing cost of service.

Strategic scheduling of residential energy consumers

We propose the strategic scheduling problem of an energy utility that administers a large residential population and may ask individual consumers to curtail part of their HVAC usage profile up to a certain effort budget, in such a way that the aggregate reductions follow a desired day-ahead goal profile. Each consumer is described by a forecast of their thermal response profile computed using a statistical model that decomposes smart meter data into a thermally-sensitive component and an intentional usage component. We propose an algorithm for computing individually-tailored optimal action schedules (e.g., automated Demand-Response control or marketing calls) for thermal energy consumption for a large sample of consumers that is based on solving a convex program. Then, we describe an approximate version of the original scheduling problem that is both interpretable and faster to compute. For this, we recast strategic scheduling as a discrete, set selection problem in which the operator is constrained by what effort structures it can request from the consumers. We propose an efficient algorithm for selecting optimal sets of consumers based on optimizing non-monotone submodular functions.