Smart Cities: Service Models, Vulnerabilities, and Resilience

IMPORTANT: The schedule has undergone some changes recently. Please take note of the updated schedule!

This workshop will take place in Room 1005 of the Osaka International Convention Center.

Abstract

Motivated by economic and technological changes, Smart Cities are being constructed upon intelligent infrastructures spanning energy, healthcare, and transportation. The advent of these societal-scale infrastructures brings with it new opportunities for improving efficiency while simultaneously exposing novel vulnerabilities. In energy societal-scale cyber-physical systems (S-CPS), for example, smart metering technologies increase the availability of streaming data thereby enabling monetization of energy savings. Such savings can be realized by employing machine learning algorithms to customize offerings to consumers. On the other hand, the availability of this fine-grained consumer/system data and the increased number of access points to the broader system expose new privacy and security risks.

The development of a S-CPS design methodology in support of resilient, sustainable operation of Smart Cities necessitates a rigorous analytical and computational framework for analyzing information exchanges between agents and for synthesizing new service models that improve efficiency. This may proceed, for instance, by introducing resilient controls for operations as well as incorporating the use of vulnerability-aware incentives to shape consumer choice, an essential element of operations.

This workshop will focus on understanding the industrial landscape, emerging service models, vulnerabilities and policy, and resilience and sustainability.

Workshop goals

Pervasive sensing yields large quantities of data. The deployment of new technologies that utilize this data has increased the need for novel service models. Such models must be privacy- and security-aware, and must consider the socioeconomic backbone of Smart Cities. This workshop is intended to gather individuals from industry and academia, including graduate students, faculty, and researchers, to discuss research in resilient, societal-scale cyber-physical systems. The aim is to understand what problems that are of practical relevance to industry and society and to identify promising directions for future research.

The talks will be broadly accessible to attendees of CDC.

Extended description

Spurred on by economic and technological changes, Smart Cities are emerging in which cheap, easily deployable sensing/actuation technologies are being integrated into everyday decision-making by the constituents of these urban areas. Smart Cities are being constructed upon intelligent infrastructures spanning energy, healthcare, and transportation. These societal-scale infrastructures are at an inflection point due to increased interdependence on new cyber-physical system (CPS) technologies such as ubiquitous mobile sensor/actuator networks, advancements in data-driven real-time learning techniques operating in the cloud, and integration of heterogeneous semi-autonomous robots into operations and management.

The advent of these societal-scale cyber-physical systems (S-CPS) brings with it new opportunities for improving efficiency while simultaneously exposing novel vulnerabilities. In energy S-CPS, for example, smart metering technologies increase the availability of streaming data thereby enabling monetization of energy savings. Such savings can be realized by employing novel machine learning algorithms to customize offerings to consumers. On the other hand, the availability of this fine-grained consumer/system data and the increased number of access points to the broader system expose new privacy and security risks. Balancing this efficiency-vulnerability tradeoff in S-CPS is key to ensuring resilience and sustainability of Smart Cities.

The development of a S-CPS design methodology in support of resilient, sustainable operation of Smart Cities necessitates a rigorous analytical and computational framework for analyzing information exchanges between agents and for synthesizing new service models that improve efficiency. Recently the US Department of Energy and NIST have issued voluntary best practices for ensuring privacy and security of the smart grid. Such policies reinforce the need for a holistic, systems-theoretic understanding of vulnerabilities such as privacy and security risks in the infrastructure systems that underly Smart Cities.

Through access to streaming data and CPS technologies that allow us to close the loop, we are seeing new service models and monetization of non-traditional commodities and goods. S-CPS models that incorporate customers into solutions in a dynamic, bi-directional way are becoming commonplace. For example, this phenomena can be seen in the rise of new car-sharing models, 3rd–party energy demand response aggregators, and the proliferation of personal health-monitoring devices. Companies are beginning to capitalize on access to consumer and system data. It is necessary to have a systematic understanding of the impact of these service models on standard operations as well as the resilience and sustainability of infrastructure systems. There is a need for tools and methods to address the efficiency-vulnerability tradeoff that is inherent to S-CPS infrastructure. This may proceed, for instance, by introducing resilient controls for operations as well as incorporating the use of vulnerability-aware incentives to shape consumer choice, an essential element of operations. Understanding consumer preferences in the presence of vulnerabilities spurred on by the utilization of new cyber-physical technologies is crucial to being able to design incentives that enable efficient operation of smart infrastructure systems.

This workshop will focus on understanding the industrial landscape, emerging service models, vulnerabilities and policy, and resilience and sustainability.

Schedule

Time Speaker Title
9:45am - 10:00am Roy Dong (University of California, Berkeley) and Lillian J. Ratliff (University of Washington) Opening remarks
10:00am - 10:30am Henrik Ohlsson (C3 Energy) C3 Energy – Analytics for Large Scale CPS
10:30am - 11:00am Sam Coogan (University of California, Los Angeles) Traffic Predictive Control: History’s Principal Components Repeat Themselves
11:00am - 11:30am Coffee break
11:30am - 12:00pm Nanpeng Yu (University of California, Riverside) Proactive Demand Participation of Smart Buildings in Smart Grid
12:00pm - 12:30pm Shuo Han (University of Pennsylvania) Privacy-Preserving Distributed Optimization and Its Application in Energy Systems
12:30pm - 2:00pm Lunch
2:00pm - 2:30pm Aron Laszka (University of California, Berkeley) Vulnerability of Transportation Networks
2:30pm - 3:00pm André Teixeira (Delft University of Technology) Attack Scenarios and Risk Metrics for Smart Grids
3:00pm - 3:30pm Andrew Clark (Worchester Polytechnic Institute) Submodularity for Scalable and Resilient Control in Smart Cities
3:30pm - 4:00pm Coffee break
4:00pm - 4:30pm Baosen Zhang (University of Washington) Understanding the Demand Side of the Grid
4:30pm - 5:00pm Roy Dong (University of California, Berkeley) and Lillian J. Ratliff (University of Washington) Roundtable discussion

Talk details

Titles and abstracts for this workshop's talks follow below.

Speaker: Henrik Ohlsson (C3 Energy) [site]

Title: C3 Energy – Analytics for Large Scale CPS

Abstract: This talk will give an industry perspective on cyber-physical systems and smart cities.

Speaker: Sam Coogan (University of California, Los Angeles) [site]

Title: Traffic Predictive Control: History’s Principal Components Repeat Themselves

Abstract: The advent of ubiquitous traffic sensing provides unprecedented real-time, high-resolution data of traffic flow that elucidate historical trends and current conditions, yet traditional signal control approaches are designed to operate with relatively limited or no real-time and/or historical data. In this talk, we propose using principle component-based decomposition techniques to learn trends in historical data that are then used to make real-time predictions of traffic flow minutes or hours into the future. We show that such predictions lead to practical, preemptive control schemes that accommodate the predicted traffic conditions rather than average traffic conditions, thus improving intersection performance and efficiency.

Speaker: Nanpeng Yu (University of California, Riverside) [site]

Title: Proactive Demand Participation of Smart Buildings in Smart Grid

Abstract: Buildings account for nearly 40% of the total energy consumption in the United States. As a critical step toward smart cities, it is essential to intelligently manage and coordinate the building operations to improve the efficiency and reliability of overall energy system. With the advent of smart meters and two-way communication systems, various energy consumptions from smart buildings can now be coordinated across the smart grid together with other energy loads and power plants. In this talk, I will present a comprehensive framework to integrate the operations of smart buildings into the energy scheduling of bulk power system through proactive building demand participation. This new scheme enables buildings to proactively express and communicate their energy consumption preferences to smart grid operators rather than passively receive and react to market signals and instructions such as time varying electricity prices. The proposed scheme is implemented in a simulation environment. The experiment results show that the proactive demand response scheme can achieve up to 10% system generation cost reduction and 20% building operation cost reduction compared with passive demand response scheme. The results also demonstrate that the system cost savings increase significantly with more flexible load installed and higher percentage of proactive customers participation level in the power network.

Speaker: Shuo Han (University of Pennsylvania) [site]

Title: Privacy-Preserving Distributed Optimization and Its Application in Energy Systems

Abstract: As many cyber-physical systems start to rely on collecting user data for more efficient operation, privacy has emerged as a concern among participating users. In the context of (coordinated) distributed decision-making, a central mediator often needs to communicate with users by exchanging messages, which may be exploited by adversaries to breach the privacy of users. In this talk, I will discuss the problem of designing privacy-preserving distributed optimization algorithms under the framework of differential privacy, which is a recently proposed notion of privacy that is robust to side information. We study a class of constrained optimization problems that are solvable using distributed gradient descent, and we show how differential privacy can be preserved by introducing additive noise in the gradients. We are also able to quantify the trade-off between the level of privacy and the loss of utility using tools from optimization theory. Finally, we demonstrate the application of the algorithm through the case of distributed charging of electric vehicles.

Speaker: Aron Laszka (University of California, Berkeley) [site]

Title: Vulnerability of Transportation Networks

Abstract: The evolution of traffic signals from standalone hardware devices to complex networked systems has provided society with many benefits, such as reducing wasted time and environmental impact. However, it has also exposed transportation networks to cyber-attacks. While traditional hardware systems were susceptible only to attacks based on direct physical access, modern systems are vulnerable to attacks through wireless interfaces or even to remote attacks through the Internet. Indeed, recent studies have shown that many traffic sensors and signals deployed in practice have easily exploitable vulnerabilities, which allow an attacker to tamper with sensor measurements or signal configurations. Due to hardware-based failsafes, these vulnerabilities cannot be used to cause accidents; however, they may be used to cause disastrous traffic congestions. In order to increase the resilience of a transportation network to such attacks, we must first assess the vulnerability of the network. Building on a well-known traffic model and a micro-model based simulator, we introduce an approach for evaluating the vulnerability of transportation networks. Besides quantifying vulnerability, the proposed approach also enables identifying controllers that have the greatest impact on congestion and which, therefore, make natural targets for attacks.

Speaker: André Teixeira (Delft University of Technology) [site]

Title: Attack Scenarios and Risk Metrics for Smart Grids

Abstract: This talk discusses some of the cyber-security challenges faced by Smart Grids. A special focus is given to the notion of “risk” and the characterization of adversarial models and threat scenarios, which are a cornerstone for performing risk analysis of such threats and to develop suitable risk mitigation approaches. Examples spanning from adversarial actions on voltage control schemes in interconnected microgrids and false-data injection attacks are discussed.

Speaker: Andrew Clark (Worchester Polytechnic Institute) [site]

Title: Submodularity for Scalable and Resilient Control in Smart Cities

Abstract: Smart city domains including energy, transportation, and health require scalable control algorithms with provable guarantees on performance, resilience, and controllability. In this talk, I will present a submodular optimization framework for satisfying these criteria. Submodularity is a diminishing returns property of set functions that leads to polynomial-time approximation algorithms for combinatorial problems. The focus of this talk will be on emerging submodular optimization-based techniques for control of energy systems.

Speaker: Baosen Zhang (University of Washington) [site]

Title: Understanding the Demand Side of the Grid

Abstract: The advance of renewable resources and distributed energy devices have dramatically changed and the form and function of the power system. In particular, instead of passive loads, the demand side is starting to actively participate in the operation of the grid. In this talk, we show how we can understand the consumption pattern of individual and a group of customers through smartmeter data. Based on these insights, we show how the traditional big utility model can be redesigned to increase the efficiency of demand side management. We also show that some long standing statistical estimation procedures are in fact inaccurate and suggest the correct estimation algorithms.