Progress Report #2

The team's progress over Spring Break 2018

Team Updates

Thi week, the team made big strides in PCB design, cloud architecture, and Lepton module testing with SPI on RPi.

Ranjini's Updates

Ranjini finished routing the PCB and is in the process of sending it out for manufacture after checking with Artur. She will place the orders for the necessary components for the PCB by the end of this week. She also explored using the ESP32 and loading freeRTOS on to it

Ash's Updates

Currently implementing GMM for density estimation for pictures. Next steps include adding to lambda in the upcoming week as well as being able to find most dense areas from a series of images rather than one image. Will be working with Arya to create multiple permissions to Lambda so that everyone can contribute code directly onto Lambda.

Arya's Updates

Arya worked on interfacing with the FLiR Lepton via SPI from a Raspberry Pi 3 Model B v1.2. The tutorial used for wiring QT package download/build, and more can be found in this Sparkfun FLiR Lepton Hookup Guide. However, Arya deviated from this demonstration to minimize the amount of code needed to debug to a more simplistic implementation of the application named raspberry_qt, after the qt tools it uses to build. The application can be found in this Github repo. The IR image captured from this exercise, demonstrating the Lepton module's functionality, is displayed to the left.

AWS Lambda PoC, S3 Bucket, CloudWatch Logs, and Billing Information

A dumby Lambda function is counting the number of black pixels are found in the sample image, which can be found in the S3 bucket image capture, named export-dcode-2018-03-15-18-51-10.txt. The CloudWatch logs contain the output from the Lambda function's successful execution.

Next Steps

Next steps for the team are to begin firmware engineering, configure AWS Lambda function permissions for multiple account access, and to begin testing occupant detection algorithm in the cloud interfacing with RPi-Lepton PoC.