Weekly Report of Team D10

March 29th 2018

Zecheng
1.Set up AWS Lambda.
Zecheng set up AWS lambda function that’s able to pass information from AWS IoT Core to our server. We were planning to run triangulation algorithm in lambda function but later decided to move that to the server. Now server is responsible for running the algorithm as well as displaying the UI.
2.Sample Application
Since we have IoT Core, Lambda function, server and UI ready, we can run a sample application where STMicrocontroller sends mock data to IoT Core and see the result of triangulation through the UI.

Bowei:
1. Build UI to display location
We will have the UI ready for Monday’s demo. It will be running on the server and users can access it through a webpage.
2. Add signal strength information to the current optimization algorithm so that more weight could be placed on sensors that have higher reliabilities.

Chen Yang:
1.Configure 2 bluecoin sensors to provide DOA output simultaneously for the demo. A host laptop is connected to the 2 sensors via usb and information is sent over uart.
2.Experiment with bluecoin sensor to understand relationship between the distance between sound source and sensor and the signal strength received at the sensor. This would be used to determine how much output from one sensor is to be trusted.

March 22nd 2018

Zecheng
1. Flash freeRTOS to STM32L4
Zecheng is done with fashing freeRTOS to STM32L4. He also set up a small application that’s able to set message directly from STM32L4 to AWS IoT Core. He will focus on retrieving sensor data from the microcontroller and the lambda service triggered by the message.

Chen:
1. Successfully set up the SMARTMIC1 demo code on BlueCoin sensor board:
Used the external ST-link programming/debug board on ST Nucleo to setup the DOA calculation demo, the result is shown by the blinking of the nearest LED to the sound source detected.
2. Test acoustic source localization performance:
Using the SMARTMIC1 demo, the acoustic source localization firmware was tested under different circumstances. Sound sources at different distances from the sensor were se tup and resulting DOA success rate were recorded.

Bowei:
1. Did simulation of four sensors by placing the only sensor at four different corners of a confined space and recorded 4 bearing readings (pretending that we actually had four sensors spitting out bearings at the same time). Using the 4 bearings estimated position of the sound source was obtained.
He will focus on reducing error by trying out different optimization methods.
Read literature on how to optimize sensor deployment in order to achieve better accuracy.

March 8th 2018

This week the STMicroController just came in and we’ve finished the setup for the microcontroller. In terms of Bluecoin sensor, we tested the directional accuracy of Bluecoin Sensor. It has a decent performance in the center of the room, but the direction is not that consistent when it is close to the wall or at the corner. We will try out different arrangements of these bluecoin sensors and probably consider placing sound-absorbing materials at the wall.

Here are the tasks we completed this week.
Chen
1.Environment Setup for STEVAL-BCNST01V1 and B-L475E-IOT01A
2.Test acoustic source localization performance
Chen has done with the environment setup for STEVAL-BCNST01V1(Bluecoin Sensor) and B-L475E-IOT01A(STMicroController). We’re now able to upload programs to the bluecoin sensor.

Zecheng
1.Create AWS IoT device registration
2.Flash freeRTOS to STM32L4 (Due on Saturday, March 10th)
Zecheng set up IoT registration on AWS. As soon as freeRTOS is flashed to the microcontroller, we’re able to send messages directly to AWS.

Bowei
1.Use bearing information from Bluecoin to do triangulation (produce absolute coordinates)
Bowe has finalized the the algorithm of triangulation and he will start implementing this algorithm next week.