Andrew L. King

PhD Student
in the Computer Information Systems Department at the University of Pennsylvania


I'm a member of the PRECISE Center and my advisor is Prof. Insup Lee

office: Levine 514


Research

Here are some projects I am involved with:

Medical Device Coordination Framework (MDCF)

A patient undergoing surgery or staying in the ICU is connected to a wide variety of medical devices. Currently, healthcare providers deliver care to the patients by manually manipulating these medical devices. If medical devices could communicate and coordinate with one another, this manipulation could be automated thereby reducing the risk of human error by reducing healthcare professional's workload, ultimately saving lives. The MDCF Project is a computational platform (middleware, a runtime execution environment, and programming abstractions) that enables medical device coordination in a systematic way. Embedded below is a demonstration of the MDCF running 2 MDCF applications: 1) A smart alarm which analyzes multiple data-streams to determine if an ICU patient needs urgent attention and 2) an application which closes the loop between a pulse-oximeter and a PCA infusion pump.


Generic Smart Alarm (GSA)

Many smart alarms and Clinical Decision Support Systems (CDSS) share architectural and algorithmic commonalities. Smart alarms and CDSS are most effective when they are targetted to a specific patient population or even specific patient. The Generic Smart Alarm (GSA) is reconfigurable processing pipeline and collection of parameterized software components designed to plug into that pipeline. Smart Alarms and CDSS built using the GSA can be quickly customized and reconfigured to target other patients and patient populations.

MIDdleware Assurance Substrate (MIDAS)

MIDAS is a distributed computational platform designed to support runtime assembly of components (e.g. subsystems containing sensors and actuators) into a larger distributed CPS. MIDAS provisions computational resources (e.g., network bandwidth and CPU time) to guarantee real-time QoS properties (e.g., all application messages reach their destinations and are processed by their deadlines). MIDAS also has a number of features designed to provide fault containment and isolation between different applications running on the same hardware or network.

Publications

2013

2012

2011

2010

2009

2006

Teaching Experience

Education

Work Experience