> INNOVATION + ENTREPRENEURSHIP
CitySensor
-
Objective:
This project aims to develop a mobile environmental monitoring platform using Arduino and research-grade air pollution and meteorology monitoring equipment. Arduino is open-source and is flexible enough to support a variety of needs for environmental sensing. So, it is being touted as a prototyping platform for Internet of Things (IoT) development. The funding support for this project comes from the USC Research Enhancement Grant.
-
Motivation:
It all came naturally from my own needs to develop a custom datalogger to connect different environmental monitoring devices for my dissertation project. Air pollution monitoring devices, such as TSI's DustTrak and Thermo Scientific's pDR 1500, all come with different communication protocols and software that requires transferring of data between the device and a computer. Often times, the transferring process takes several hours in addition to the time it takes to collect samples in the field.
-
Components:
-
DataLogger: Arduino YUN and Arduin Uno
-
TSI DustTrak 8520: capable of measuring PM 10 or PM 2.5 concentrations in outdoor environment. The communication protocol is RS232, and uses specific command to communicate with the device.
-
TSI CPC 3007: capable of measuring much finer ultrafine particle concentrations than the P-trak. It operates with the same principle as the P-Trak, but it has faster response to freshly emitted traffic pollution, therefore, it is used widely for measuring near-road traffic pollution.
-
TSI P-Trak: capable of counting ultrafine particles in outdoor environment. It operates by drawing an aerosol sample continuously through a heated saturator, where alcohol is vaporized and diffuses into the sample stream. Isopropyl alcohol cartridge needs to be replaced every 3 hours depending on the temperature of a sampling site.
-
Wind direction/speed sensor: Gill WindSonic Ultrasonic Anemometer
-
Temperature / Relative humidity sensor: Vaisala HMP50 temperature and humidity probe
-
What is Arduino?
Arduino is a small but powerful microcontroller (i.e. computer) that allows users to interface with a variety of electronic devices and sensors. It was originally developed by Massimo Banzi and David Cuartielles to teach their students about electronic circuit boards, but now its use has been expanded beyond the classroom.
Arduino is now being used and implemented by commercial developers and engineers as a cross platform to connect and control different devices. The beauty of Arduino is that, thanks to its open-source approach, users can customize any electronic devices to suit your needs with simple lines of codes based on C programming language. There is also a huge community of developers, hobbyists, students, and researchers that help each other's projects.
-
Basic Architecture:
The basic architecture of the CitySensor follows a typical Internet of Things (IoT) framework. Basically, a set of sensors are connected to the central microprocessor (Arduino) through either digital channel or analog channel (Data Collection). The collected sensor data are being processed and formatted using a timestamp through the centralized microprocessor (Data Processing). The microprocessor has a 3G network communication extension which allows the processed data to be transmitted over cellular network (Data Transmission). The transmitted data are being archived on a cloud server that runs 24 hours and 365 days (Data Storage). Users can access both the archived data and the data being streamed through the cellular network in real time.
-
Prototype Development:
A prototype was developed using Arduino as a central microprocessor. The prototype successfully read both analog and digital signals. One major challenge of this project is that different devices use different communication protocols. So, it is difficult to directly connect monitoring devices to the Arduino system, and expect Arduino to automatically read the signals. Analog devices are much simpler because there are no protocol issues. But when it comes to reading digital signals, each manufacturer and even with the same manufacturer, different models use different protocols. Currently, the prototype is under development and in a testing phase. For field deployment, a weather-proof enclosure was designed. Several enclosures were considered, including water-proof NEMA enclosure and Pelican brief enclosure. NEMA 3R was chosen as the final enclosure type because of it's sturdy design and weather resistant function. See this document for the design of the NEMA 3R-based environmental enclosure.
 Arduino + DustTrakDustTrak was connected to Arduino. Arduino reads digital signals from the DustTrak and save the data stream into an SD card. |  Arduino and breadboardThis is a basic Arduino and breadboard setup. |
|---|---|
 RS232 connectionConnection between Arduino and DustTrak through RS232 protocol. |  Arduino + DustTrak + WindSonicThis is the final configuration of Arduino with DustTrak, temperature/humidity sensor, and wind anemometer connected together. |
 TSI DustTrak 8520 |  Gill WindSonic Anemometer |
 Vaisala HMP50 temp/humidity sensor |  Arduino LabI turned the urban planning office into a messy computer engineering lab. People have no clue about what I'm doing here. |
 CitySensor Prototype 1This is the very first prototype of CitySensor. It is a plain waterproof NEMA 3R enclosure. |  CitySensor Prototype 2This is a second prototype of CitySensor. A shelf and air vent was added to the first version. It also has handles and a cane pipe that functions as an inlet. It has the 6 feet pipe to which anemometer will be attached. |
 CitySensor Prototype 3This is a slightly improved version three. The cane pipe was replaced with a cone pipe, and the enclosure was lifted using four legs. |  CitySensor Prototype 4This is the final version ready to be manufactured. The legs were shorted to improve stability. |
-
Prototype Deployment:
We had a series of discussion with Los Angeles Metro to pilot test the CitySensor by deploying it at various transit stations in Los Angeles. Two options have been suggested. One option was to put the device in a metal cage to provide full protection. Another option was to fix the equipment using a metal chain. For either options, deployment must be coordinated with union workers and transit police. For added security, it was suggested that the Metro place a CCTV camera to strengthen security level near the deployment area. The photos below show site visits and feasibility testing of the suggested deployment options.
Option 1: Full Protection using a Metal Cage
Option 2: Partial Protection using a Metal Chain
Feasibility Assessment for Prototype Deployment at various Los Angeles Metro Stations
BikePed Analytics
-
Objective:
This project was born out of a collaboration with UCLA and Placemeter (TM) to validate computer-vision technology for quantifying traffic information, such as pedestrian and bicycle movement, and passenger and freight vehicles.
-
Motivation:
The motivation to partner with Placemeter came while I was doing a study on the impact of CicLAvia (car-free street event) on traffic and air quality. Traditionally, we have been using pneumatic tubes and human labor to count pedestrians, bicyclists, and cars. In liue of this old technology, Placemeter offers much easier and cheaper ways to quantify traffic data which do not require any equipment installation or human labor. If this approach can be proven successful for quantifying bike/peds with higher precision, this technology has the potential to disrupt the traditional traffic engineering practice of quantifying bikes/peds/vehicles.
-
Progress:
-
One of the major challenges of this technology is to get access to city-owned CCTVs or IP cameras. New York City already opens their traffic CCTVs to the public, so Placemeter has no problem getting the traffic feeds from the city. For the Culver City CicLAvia event, we successfully recorded traffic camera feeds. Special thanks to Gabe Garcia (traffic engineering Analyst at Culver City) who helped us get the traffic camera feeds. We have successfully turned the traffic videos into valuable traffic count data.
In the case of Los Angeles, however, there is no publicly available CCTV camera feeds, so we had to get permission from LA Department of Transportation to get access to their traffic feeds. It turns out that they have legal and security issues with regard to plugging any device to their system. So, we couldn't use Placemeter technology for our project. In the future, if many cities adopt internet-connected CCTV cameras and made them available publicly, Placemeter technology has a bright future. At the moment, the legal and security hurdles in getting access to city-owned traffic feeds are the major barriers to this new technology.
What is Placemeter?
Placemeter is a company based in New York City that develops computer vision algorithm to offer users with convenient ways to quantify foot, bike, and vehicle traffic. They process video feeds from CCTV cameras, and their algorithm detects pedestrians, bicyclists, and vehicles with an accuracy of close to 90%.
The computer vision algorithm was developed a long time ago around 60s, but what's unique about Placemeter is that they provide very convenient front-end user interface to monitor traffic in realtime. They also allow users to record video feeds using their smart phones and use that video to quantify traffic information.