1 Remote Sensing Jordan Adelman, Dan Reicher, Norissa Lamaute, Dallas van Norden, Matt Ganis

2 Goal The purpose of this project was to create an underwater sensor system that Collects data about the environment (i.e. temperature, oxygen level, etc.) Saves the data to a database Displays the data for our researchers in a comprehensive and effective format

3 Scalability

4 Architecture Adafruit Feather Pipe Design Battery/Solar Power

5 Removable Top/bottom for access to sensor connectionsSensors embedded into the “pipe”

6 Monitoring Functions Monitor battery level (Are we charging)Moisture in the case microcontroller battery Network: Are we sending data Are the cloud applications ”up” Water sensor inside the pipe to see if we’re taking in water Application monitoring

7 The Action Via Pace Wi-Fi Measure the water temperature of Choate Pond at three distinct levels and compare the changing temperature to the current air temperature

8 Cloud Computing using Customer has convenient access to IBM Bluemix.Bluemix is IBM’s cloud platform that combines PaaS with IaaS and offers a catalog of cloud services. IBM Internet of Things service provides an API to connect devices to Bluemix’s IoT cloud. For an agile deployment, we took advantage of IBM’s cloud platform, IBM Bluemix. This would allow us to scale data processing with little time-consuming resource management. [IBM 2017] [https://console.ng.bluemix.net/catalog/services/internet-of-things-platform/] [https://www.ibm.com/blogs/nordic-msp/wp-content/uploads/2016/09/Bluemix-300x111.png]

9 Cloud Security Privacy, access control, and identity managementMany companies use cloud computing providers for data operations now Business continuity and data integrity must be addressed before changes

10 Managing Big Data with Problem: Too many data pointsSolution: Scalable NoSQL can fit the millions of data points MongoDB is a NoSQL DBMS and is readily available in IBM Bluemix. Problem: If temperature data is collected uninterrupted from 3 sensors every minute for an entire year, there would be over 1.5 million data points. The team used a MongoDB database to organize the collected data. There is one database for each sensor, which in our first prototype includes one sensor/database for each level of the pond. The tables are organized by the date, and inside of each table, the collection is organized first by the hour from 00-23, then by the minute from , and finally by the temperature (measured in Fahrenheit) from So that a sample temperature might appear to be in database 2, table , and collection This would tell us that for sensor 2 on March 5, 2017, the temperature was 63 degrees Fahrenheit at 5:48 AM.

11 MQTT Bare-Bones Not resource-intensive Very low security Fast and easy to use Publish/Subscribe protocol for receiving data sent by sensor

12 Node-RED A browser-based flow editor to "wire" together devices, APIs and web services An organizational framework that visually and functionally represents the flow of data

13

14 Node-RED Platform

15 Results Live results available at: http://env-sense-mrg. mybluemix

16 What we’ve done so far… Researched and implemented the tools in this prototype Created the code/physical prototype that collects and inputs data Designed the database Developed a simple way to visualize the collected data

17 Future Goals Test water for Oxygen content, Nitrogen content, possibly more Develop a more robust visualization technique Collect data from Hudson River with a more scalable design

18 Acknowledgements and ConclusionDr. Ganis, Mackenzie Dolishny, Jennifer Gillen, Lubin School of Business, Dyson, Seidenberg School of CSIS

19