The objective of this research is to develop a theory of ActionWebs, that is, networked embedded sensor-rich systems, which are taskable for coordination of multiple decision-makers. The approach is to first identify models of ActionWebs using stochastic hybrid systems, an interlinking of continuous dynamical physical models with discrete state representations of interconnection and computation. Second, algorithms will be designed for tasking individual sensors, based on information objectives for the entire system. Third, algorithms for ActionWebs will be developed using multi-objective control methods for meeting safety and efficiency objectives. Two grand challenge applications for this research are in Intelligent Buildings for optimal heating, ventilation, air conditioning, and lighting based on occupant behavior and external environment; and Air Traffic Control for mobile vehicle platforms with sensor suites for environmental sensing to enable safe, convenient, and energy efficient routing. The intellectual merit of this research stems from a conceptual shift of ActionWebs away from passive information gathering to an action-orientation. This involves: modeling of ActionWebs using stochastic hybrid systems; taskable, multi-modal, and mobile sensor webs; and multi-scale action-perception hierarchies. The broader impact of the research is in two grand challenge national problems: energy efficient air transportation, and energy efficient, high productivity buildings, and will tackle social, privacy, economic, and usability issues. Integrated with the research is a program of coursework development in networked embedded systems, across stove pipes in EECS, Aero-Astro, Civil, and Mechanical Engineering departments. Outreach objectives include new course design at San Jose State University, and recruiting more women researchers.

The objective of this research is to develop new principles for creating and comparing models of skilled human activities, and to apply those models to systems for teaching, training and assistance of humans performing these activities. The models investigated will include both hybrid systems and language-based models. The research will focus on modeling surgical manipulations during robotic minimally invasive surgery. Models for expert performance of surgical tasks will be derived from recorded motion and video data. Student data will be compared with these expert models, and both physical guidance and information display methods will be developed to provide feedback to the student based on the expert model. The intellectual merit of this work lies in the development of a new set of mathematical tools for modeling human skilled activity. These tools will provide new insights into the relationship between skill, style, and content in human motion. Additional intellectual merit lies in the connection of hybrid systems modeling to language models, the creation of techniques for automated training, and in the assessment of new training methods. The broader impact of this research will be the creation of automated methods for modeling and teaching skilled human motion. These methods will have enormous implications for the training and re-training of the US workforce. This project will also impact many diversity and outreach activities, including REU programs and summer camps for K-12 outreach. The senior personnel of this project also participate in the Robotic Systems Challenge and the Women in Science and Engineering program.

The objective of this research is to develop principles and tools for the design of control systems using highly distributed, but slow, computational elements. The approach of this research is to build an architecture that uses highly parallelized, simple computational elements incorporating nonlinearities, time delay and asynchronous computation as integral design elements. Tools for the design of non-deterministic protocols will be developed and demonstrated using an existing multi-vehicle testbed at Caltech. The motivation for using "slow computing" is to develop new feedback control architectures for applications where computational power is extremely limited. Examples of such systems are those where the energy usage of the system must remain small, either due to the source of power available (e.g. batteries or solar cells) or the physical size of the device (e.g. microscale and nanoscale robots). A longer term application area is in the design of control systems using novel computing substrates, such as biological circuits. A critical element in both cases is the tight coupling between the dynamics of the underlying process and the temporal properties of the algorithm that is controlling it. The implementation plan for this project involves students from multiple disciplines (including bioengineering, computer science, electrical engineering and mechanical engineering) as well as at multiple experience levels (sophomores through PhD students) working together on a set of interlinked research problems. The project is centered in the Control and Dynamical Systems department at Caltech, which has a strong record of recruiting women and underrepresented minority students into its programs.

The objective of this research is to develop an integrated methodology for control system design in situations where disturbances primarily result from routine human behavior, as, for example, in future artificial pancreas systems where meals and exercise are the main disturbances affecting blood glucose concentration. The approach is to recognize that human behavioral disturbances (i) are generally random but cannot be treated as zero-mean white noise processes and (ii) occur with statistical regularity but cannot be treated as periodic due to natural variation in human behavior. This emerging class of problems requires (i) the derivation of new mathematical representations of disturbances for specific applications and (ii) the formulation of new stochastic control models and algorithms that exploit statistical regularity in the disturbance process. The intellectual merit of the proposed research stems from the fact that it explicitly recognizes a new class of disturbances, human behavioral disturbances, seeking to develop an integrated approach to statistically characterizing and responding to future perturbations, adapting gracefully to uncertainty about the future. The anticipated research outcomes will be relevant in diverse fields, including stochastic hybrid control and human automation interaction. As a broader implication, the proposed research will enable the design of future field deployable artificial pancreas systems, potentially improving the lives of 1.5 million Americans suffering from Type 1 diabetes. With help from the two graduate students funded by the project, the principle investigator will supervise a Capstone design course, exposing undergraduates to various aspects of control under human behavioral disturbances.

This objective of this proposal is to improve the management of the air traffic system, a cyber-physical system where the need for a tight connection between the computational algorithms and the physical system is critical to safe, reliable and efficient performance. The approach is based on an adaptive multiagent coordination algorithm with a particular emphasis on the systematic selection of the agents, their actions and the agents' reward functions. The intellectual merit lies in addressing the agent coordination problem in a physical setting by shifting the focus from "how to learn" to "what to learn." This paradigm shift allows a separation between the learning algorithms used by agents, and the reward functions used to tie those learning systems into system performance. By exploring agent reward functions that implicitly model agent interactions based on feedback from the real world, this work aims to build cyber-physical systems where an agent that learns to optimize its own reward leads to the optimization of the system objective function. The broader impact is in providing new air traffic flow management algorithms that will significantly reduce air traffic congestion. The potential impact cannot only be measured in currency ($41B loss in 2007) but in terms of improved experience by all travelers, providing a significant benefit to society. In addition, the PIs will use this project to train graduate and undergraduate students (i) by developing new courses in multiagent learning for transportation systems; and (ii) by providing summer internship opportunities at NASA Ames Research Center.

The objective of this research is to develop technologies to improve the efficiency and safety of the road transportation infrastructure. The approach is to develop location-based vehicular services combining on-board automotive computers, in-car devices, mobile phones, and roadside monitoring/surveillance systems. The resulting vehicular Cyber Physical Systems (CPS) can reduce travel times with smart routing, save fuel and reduce carbon emissions by determining greener routes and commute times, improve safety by detecting road hazards, change driving behavior using smart tolling, and enable measurement-based insurance plans that incentivize good driving. This research develops distributed algorithms for predictive travel delay modeling, feedback-based routing, and road hazard assessment. It develops privacy-preserving protocols for capturing and analyzing data and using it for tasks such as congestion-aware tolling. It also develops a secure macro-tasking software run-time substrate to ensure that algorithms can be programmed centrally without explicitly programming each node separately, while ensuring that it is safe to run third-party code. The research focuses on re-usable methods that can benefit multiple vehicular services, and investigates which lessons learned from this vehicular CPS effort generalize to other situations. Road transportation is a grand challenge problem for modern society, which this research can help overcome. Automobile vendors, component developers, and municipal authorities have all shown interest in deployment. The education plan includes outreach to local K-12 students and a new undergraduate course on transportation from a CPS perspective, which will involve term projects using the data collected in the project

The objective of this research is to create computational foundation, methods, and tools for efficient and autonomous optical micromanipulation using microsphere ensembles as grippers. The envisioned system will utilize a holographic optical tweezer, which uses multiple focused optical traps to position microspheres in three-dimensional space. The proposed approach will focus on the following areas. First, it will provide an experimentally validated optical-tweezers based workstation for concurrent manipulation of multiple cells. Second, it will provide algorithms for on-line monitoring of workspace to support autonomous manipulation. Finally, it will provide real-time image-guided motion planning strategies for transporting microspheres ensembles. The proposed work will lead to a new way of autonomously manipulating difficult-to-trap or sensitive objects using microspheres ensembles as reconfigurable grippers. The proposed work will also lead to fundamental advances in several cyber physical systems areas by providing new approaches to micromanipulations, fast and accurate algorithms with known uncertainty bounds for on-line monitoring of moving microscale objects, and real-time motion planning algorithms to transport particle ensembles. The ability to quickly and accurately manipulate individual cells with minimal training will enable researchers to conduct basic research at the cellular scale. Control over cell-cell interactions will enable unprecedented insights into cell signaling pathways and open up new avenues for medical diagnosis and treatment. The proposed integration of research with education will train students with a strong background in emerging robotics technologies and the inner workings of cells. These students will be in a unique position to rapidly develop and deploy specialized robotics technologies.

The objective of this research is to enable operation of synthetic and cyborg insects in complicated environments, such as outdoors or in a collapsed building. As the mobile platforms and environment have significant uncertainty, learning and adaptation capabilities are critical. The approach consists of three main thrusts to enable the desired learning and adaptation: (i) Development of algorithms to efficiently learn optimal control policies and dynamics models through sharing the learning and adaptation between various instantiations of platforms and environments. (ii) Creation of control learning algorithms which can be run on low-cost, low-power mobile platforms. (iii) Development of algorithms for online improvement of policy performance in a minimal number of real-world trials. The proposed research will advance learning and adaptation capabilities of practical cyberphysical systems. The proposed approach will be generally applicable and lead to a new class of learning and adapting systems that are able to leverage shared properties between multiple tasks to significantly speed up learning and adaptation. Success in this research project will bring society closer to solving the grand challenge of teams of mobile, disposable, search and rescue robots which can robustly locomote through uncertain and novel environments, finding survivors in disaster situations, while removing risk from rescuers. This project will provide interdisciplinary training through research and classwork for undergraduate and graduate students in creating systems which intimately couple the cyber and physical aspects in robotic and living mobile platforms. Through the SUPERB summer program, under-represented students in engineering will experience research in learning and robotics.

The objective of this research is to study the formal design and verification of advanced vehicle dynamics control systems. The approach is to consider the vehicle-driver-road system as a cyber-physical system (CPS) by focusing on three critical components: (i) the tire-road interaction; (ii) the driver-vehicle interaction; and (iii) the controller design and validation. Methods for quantifying and estimating the uncertainty of the road friction coefficient by using self-powered wireless sensors embedded in the tire are developed for considering tire-road interaction. Tools for real-time identification of nominal driver behavior and uncertainty bounds by using in-vehicle cameras and body wireless sensors are developed for considering driver-vehicle interaction. A predictive hybrid supervisory control scheme will guarantee that the vehicle performs safely for all possible uncertainty levels. In particular, for controller design and validation, the CPS autonomy level is continuously adapted as a function of human and environment conditions and their uncertainty bounds quantified by considering tire-road and driver-vehicle interaction. High confidence is critical in all human operated and supervised cyber-physical systems. These include environmental monitoring, telesurgery, power networks, and any transportation CPS. When human and environment uncertainty bounds can be predicted, safety can be robustly guaranteed by a proper controller design and validation. This avoids lengthy and expensive trial and error design procedures and drastically increases their confidence level. Graduate, undergraduate and underrepresented engineering students benefit from this project through classroom instruction, involvement in the research and substantial interaction with industrial partners from the fields of tires, vehicle active safety, and wireless sensors.

The objective of this research is the transformation from static sensing into mobile, actuated sensing in dynamic environments, with a focus on sensing in tidally forced rivers. The approach is to develop inverse modeling techniques to sense the environment, coordination algorithms to distribute sensors spatially, and software that uses the sensed environmental data to enable these coordination algorithms to adapt to new sensed conditions. This work relies on the concurrent sensing of the environment and actuation of those sensors based on sensed data. Sensing the environment is approached as a two-layer optimization problem. Since mobile sensors in dynamic environments may move even when not actuated, sensor coordination and actuation algorithms must maintain connectivity for the sensors while ensuring those sensors are appropriately located. The algorithms and software developed consider the time scales of the sensed environment, as well as the motion capabilities of the mobile sensors. This closes the loop from sensing of the environment to actuation of the devices that perform that sensing. This work is addresses a challenging problem: the management of clean water resources. Tidally forced rivers are critical elements in the water supply for millions of Californians. By involving students from underrepresented groups, this research provides a valuable opportunity for students to develop an interest in engineering and to learn first hand about the role of science and engineering in addressing environmental issues.