Seminars

This talk will focus on the design of distributed estimation systems that are formed by multiple non-collocated components. A shared network is used to disseminate information among the components.

I will discuss two recent results: Assuming that the network is characterized by an incomplete directed communication graph, the first result characterizes the existence of omniscience-achieving schemes for which all components that observe only a portion of the output of an underlying plant can estimate the entire state with error that vanishes asymptotically. Our approach hinges on key…

In the age of cloud-connected mobile devices, users want responsive apps that read and write shared data everywhere, at all times, even if network connections are slow or unavailable. Replication and eventual consistency, while able to deliver this experience, require us to face the complexity of asynchronous update propagation and conflict resolution. Our research goal is to find abstractions that encapsulate this complexity, in order to simplify the programming of distributed applications that are responsive, reactive, and collaborative.

In this talk, we first discuss the general…

Increased congestion on the airport surface has increased taxi times, fuel consumption, and emissions. In this talk, I will describe how operational data can be used to develop and validate queuing models of airport operations. These models yield new insights on the effect of different factors such as weather, runway usage, and aircraft fleet mix on airport performance. They also help us predict the behavior of an airport under different operating conditions. 

I will then show that these queuing models can be used to design Pushback Rate Control, a new airport congestion control…

Industrial-scale embedded control systems are often developed in the model-based development (MBD) paradigm. The design process typically involves capturing a 'plant model' to describe the physical processes to be controlled, and a 'controller model,' which is usually a block-diagram-based representation of the controller software. In practice, plant models and controller models are highly complex; typical plant models feature highly nonlinear and hybrid differential and algebraic equations, while controller models feature large look-up tables storing pre-computed values, several levels of…

It used to be that we built electronic systems to interact with specific users: taking inputs, providing outputs. Seeing the incredible value that electronics has brought to so many applications, today we would like to build system that interact with all aspects of the physical world. But, the physical world is complex, both in the number of relevant signals it presents and in the physics of how these signals encode information. This talk starts by exploring algorithmic frameworks for creating high-quality models for analyzing sensor signals when no tractable analytical models are…