Meghan Clark

Meghan Clark

Postdoctoral Scholar

University of California, Berkeley

Biography

I have ten years of experience developing end-to-end solutions at the intersection of IoT, mixed reality, and AI. I recently completed my PhD in computer science from UC Berkeley, where I focused on smart buildings. I’m interested in distributing computation across cloud-edge architectures, particularly to support IoT, networking, and 5G applications.

I am currently looking for positions!

Download my resumé.

Interests
  • Network monitoring
  • Sensor networks/IoT
  • Microservice architectures
  • Observability
Education
  • PhD in Computer Science, 2021

    University of California, Berkeley

  • MS in Computer Science and Engineering, 2017

    University of Michigan

  • BS in Computer Science, 2011

    George Mason University

Projects

*
Mixed Reality Network Introspection
I developed a network introspection system and mixed reality visualization for viewing communication in sensor-gateway-cloud IoT architectures. Technologies included OpenThread, WiFi, ultra-wideband localization, Unity, AWS, MQTT, Python, Golang, C#. Currently incorporating spectrum sensing.
Mixed Reality Network Introspection
Vehicular Network Visualization
In a collaboration with IBM Research, I streamed network capture, telemetry, and camera feeds from a self-driving car simulator into Unity to visualize autonomous vehicle communication during collaborative perception tasks.
Vehicular Network Visualization
Augmented Reality IoT App
I designed and built an augmented reality Android app to help users discover smart home capabilities and have successful interactions with intelligent assistants in unfamiliar smart spaces. I also led user research and evaluations.
Augmented Reality IoT App
Open Source IoT Library
I wrote and maintain lifxlan, an open source Python library for communicating with LIFX brand WiFi smart bulbs over a local network using the UDP LAN protocol. 450+ stars on Github.
Open Source IoT Library
Power Meter Analytics
I used machine learning to calibrate readings from energy harvesting power meters.
Power Meter Analytics
Smart Home Telepresence
I prototyped a smart home telepresence application that uses technology present in smart homes today. I developed a lightweight microservice runtime that allowed the same application code to run in smart homes with different device brands and configurations.
Smart Home Telepresence
Imitation Learning for Smart Lighting Control
I implemented an LSTM deep learning recurrent neural network (RNN) using the Python Keras framework. I trained the system to anticipate when home residents will turn on lights and do it for them automatically.
Imitation Learning for Smart Lighting Control
Design and Fabrication
I designed and fabricated a CubeSat-themed mood light powered by Arduino. This required PCB design and fabrication, aluminum machining (vertical mill, bandsaw), and laser cutting.
Design and Fabrication
CS Outreach Program
I founded the CS Kickstart summer program at University of Michigan to increase recruitment and retention of women in computer science. I assembled and led a five-person leadership team that built the program from the ground up. Our program piloted interventions that were eventually rolled into the introductory CS curriculum.
CS Outreach Program

Publications

Quickly discover relevant content by filtering publications.
(2021). XRShark: Mixed Reality Network Introspection. White Paper.

PDF Cite Project

(2018). Bluetooth Low Energy in the Wild Dataset. DATA ‘18.

PDF Cite Publisher

(2018). The Big House Dataset: Desired Applications and Interactions. DATA ‘18.

PDF Cite Publisher

(2018). Experience: Android Resists Liberation from Its Primary Use Case. MobiCom ‘18.

PDF Cite Publisher

(2016). Perpetual Sensing for the Built Environment. IEEE Pervasive Computing.

PDF Cite Publisher

(2014). Deltaflow: Submetering by Synthesizing Uncalibrated Pulse Sensor Streams. ACM e-Energy ‘14.

PDF Cite Publisher

(2013). Mission-Aware Cyber-Physical Optimization on a Tabletop Satellite. I@A.

PDF Cite Publisher