Charlie Mydlarz | Research Assistant Professor, Brooklyn

Charlie Mydlarz
Research Assistant Professor
Brooklyn

I’m an acoustician/engineer who designs, develops, and deploys advanced IoT devices to improve our understanding of: spaces, machinery health, and urban conditions.

Profile

I design, implement, and deliver end-to-end systems to tackle different challenges, including: city-scale urban noise sensing, acoustic condition monitoring, urban flood detection, global soundscape perception, building/classroom efficiency, and urban mobility. In developing these solutions I’ve mentored teams of scientists from undergraduate to postdoc level and successfully collaborated with stakeholders and partners such as: city agency personnel, urban residents, product development agencies, media agencies, and electrical fabrication houses. I’m able to deliver on projects requiring a high level of technical expertise and management.

I’m currently at the NYU Center for Urban Science and Progress (CUSP) and the Music and Audio Research Lab (MARL).

Projects

SONYC

I lead the hardware and systems team on the Sounds Of New York City or SONYC project. I designed, fabricated and deployed a network of over 100 low-cost acoustic sensors across NYC, many of which have been running for over four years, to provide continuous, real-time, and accurate monitoring of urban noise. I designed and supervised the building of the back and front end systems for a scalable and resilient solution. This has allowed me to carry out large-scale analysis of urban noise activity to reveal patterns across space and time. These platforms have been used to create a data-driven approach for city agencies tasked with noise mitigation by feeding relevant, actionable, and timely noise information.

REIP

The Reconfigurable Environmental Intelligence Platform (REIP) significantly enhances the capabilities of remote sensor networks. I lead the hardware and systems design of this novel sensor system that makes use of a set of hardware modules that connect together to form an advanced sensing solution. Each sensing module will come in a number of variants allowing the end user to find the proper tradeoff between complexity/cost and power/features. The system uses a reconfigurable combination of: high resolution cameras, a 15 channel microphone array, environmental sensors, and cutting edge machine learning leveraging sensor fusion to deliver insight on environments for various applications. Our test cases include: indoor sensing for reducing waste in HVAC (heating, ventilating, and air conditioning) systems, and the monitoring of urban road/pedestrian traffic. The system has the ability to bring down remote sensor network deployment times from years to weeks.

DeepSense

I was co-founder of a startup concept that aimed to provide cutting edge ultra-wideband acoustic condition monitoring solutions to manufacturing machinery OEMs who rely on continuous mechanical operation and timely, actionable foresight of impending machinery faults. Early stage machinery malfunction can be indicated by abnormal acoustic emissions, such as those produced by bearing faults, slipping belts, or loose chains. Reactive, late-stage fault detection leading to machinery failure results in high parts/repair costs, and extensive losses due to longer periods of unscheduled downtime. Predictive and data-driven, early-stage fault detection allows for strategic planning of machinery repairs including more efficient use of scheduled downtime periods resulting in increased overall asset uptime. We made use of our experience in acoustic remote sensing, digital signal processing (DSP), and machine learning to enable us to extract insight from very noisy signals.

FloodSense

I lead the sensing research and development on this USDOT and ESD funded project developing novel sensing solutions for realtime urban flood monitoring. The project is partnered with a number of other institutions and city agencies with our group’s website here (floodnet.nyc) In New York City, sea-level rise has led to a dramatic increase in flood risk, particularly in low-lying and coastal neighborhoods. High-intensity rain storms compound this risk by conveying large volumes of water to drains, leading to backups and overflows. The resulting presence of standing water on streets and sidewalks can impede mobility and restrict access to transportation. Urban flood water can also contain a diverse array of contaminants, including fuels, raw sewage, and industrial/household chemicals. Access to real-time information on flooding can improve resiliency and efficiency by allowing residents to identify navigable transportation routes and make informed decisions to avoid exposure to flood water contaminants. Flood data can also aid city agencies in delineating areas to target for flood control improvements and making data-driven decisions on flood management. However, very little data exist on the frequency and extent of urban surface flooding, and there is an unmet need for hyperlocal information on the presence and depth of street-level flood water.

Sound Around You

My citizen science driven PhD project where people across the world were prompted to use their smartphone or audio recorder to capture clips from different sound environments, or soundscapes. These clips were uploaded to our online map, along with people’s opinions of them and why they chose to record it. My thesis investigated how sounds in our everyday environment make us feel. I designed and built the full-stack system including all web content and native apps.

Skills

Smart Cities IoT
• Remote sensing
• Big data collection, processing and analysis
• Electrical engineering including: small-scale SMT PCB design, bring-up, and reworking
• Resilient hardware design and fabrication
• Full stack baremetal Linux systems engineering, admin, and dev-ops
• Networking technologies and implementations
• City agency and partner liaison
• Report creation and dissemination to stakeholders
• Trying not to complain about the cold

Acoustics
• Laboratory and field measurement
• Digital signal processing
• Large scale acoustic data processing and analysis
• Embedded audio system design
• MEMS microphones
• Structure-borne sound measurement
• Calibration techniques
• Acoustic ecology

Research
• Research project conception, development and management
• Academic publishing in peer reviewed conferences and journals
• Grant proposal writing and budgeting
• Mentoring of over 50 students from high school to postdoctoral levels

Management and planning
• Planning, organization, and delivery of 8 projects of varying sizes and team compositions
• Grant proposal writing on 6 successfully funded National Science Foundation projects totaling over $8M
• Conflict resolution in various team scenarios

Entrepreneurial
• Customer discovery
• Business model creation and analysis
• Development of evidence-based commercialization strategies

Tools and languages
• Python: num/scipy, pandas, sklearn, librosa, tornado, multiprocessing, matplotlib, jupyter notebooks
• Ansible
• Docker, including: kubernetes and compose
• Shell scripting
• Objective C, swift and android java
• C, C++, C#
• Javascript, PHP, HTML, CSS
• Matlab
• Fusion 360
• Eagle, KiCad
• Photoshop, Illustrator, Premiere, InDesign, Audition

Positions

Oct 2013 – Today

New York University

Research Assistant Professor, Senior Research Scientist, Postdoctoral Associate

Working on the development and implementation of cyberphysical systems for large scale, high resolution data capture from urban environments.

Jan 2013 – Jun 2013

University of Salford

Research Assistant

Framework for Innovation and Research in MediaCityUK (FIRM), Hyperlocal TV research project investigating hyperlocal TV services, with a particular focus on sports coverage. The project is partnered with BT, and a number of local charities and media organisations. The main aim is to gauge audience appreciation towards hyperlocal content and the ways in which they consume it.

Feb 2010 – Jul 2012

University of Salford

Research Technician

EPSRC funded project ISESS (Identifying a Sound Environment for Secondary Schools): a three year research project investigating the effects on teaching and learning of different acoustic designs within secondary schools and classrooms.

Sep 2007 – Feb 2013

University of Salford

PhD Student

Enabling public engagement in a large-scale mass participation soundscape study utilizing smart phones for global subjective and objective data collection and analysis.

Sep 2005 – Sep 2006

HHB Professional Audio Products

Service Engineer

In charge of the service and repair of professional audio equipment, including all microphones and CD burners. Ran all of the companies pro audio CD media testing and handled phone and email technical support for industry professionals.