jtfinlay/burgess

Wireless monitoring system designed to track customers through retail space.

burgess's Issues

Task 6.2 - Simple Android app that allows for calibration data generation

Task 4.4 - Send data for processing and receive position information

Dependent upon (5.1).

Task 7.2 - Gather data from server with current customer/employee locations

Depends upon (4.5) but may be mocked in

System Integration

Ensure all my values are properly set for presentations.

Task 4.2 - Receive metadata from Bluetooth application

May be mocked in to avoid dependency upon (2.5)

Task 2.3 - Implement Bluetooth metadata collection loop at required intervals

Create full day simulator

The full day simulator will simulate a full day of people entering, being helped and leaving the 'store'. The purpose of this is to provide some meat for James' analytics to crunch and to look nice when presented. The idea is to create something that appears to be an actual full day with few customers in the morning and evening, a spike mid day and a drop off around lunch. Employee's to customers will not be purposefully matched so that some parts of the day show poor service while others show good service.

All BT gathering in Android App as a seperate .jar

We said the api would be imported and integrated into other apps. Make sure the code is actually packaged seperately and then imported. @Tyler-Meen Assigning this to you to take care of.

Task 3.2 - Create loop to generate pulse. Likely built in, but may need a machine running a script

Task 7.1 - Setup Employee Android app to look like web view

Task 5.1 - Implement FSPL equation

Algorithm / Library that accepts raw metadata and perform naïve distance calculations using the Free Space Path Loss (FSPL) equation. These will not be completely accurate but should produce a workable result in a clear space.

Create test cases for WiFi Solver

The WiFi solver is not actually being used by our implementation due to conflicts between the theoretical data expected and the highly variable data actually observed. However, in theory the solver should work well if the captured data was of a higher quality. Create test cases that simulate this higher quality data. Calculations will need to be done in reverse by hand... so this will be rather time consuming.

Task 2.1 - Evaluate Bluetooth auto connect to beacons, implement if useful

Task 8.2 - Design database and setup access

Task 9.3 - Implement template layout

Task 5.2 - Algorithm to determine when employees are close to customers

Task 4.5 - Design REST API

4.5 Design REST API for querying the information. Should be able to allow for queries of raw data, position data, Employee information, and Customer categorization. These should be filterable based upon time and other data specific filters that are deemed reasonable.

Task 4.6 - Store employee data

Task 4.1 - Receive signal metadata from a tier 1 WiFi system

May be mocked in to test and eliminate dependency on (1.3)

Cross-browser compatibility

Make sure UI works on newest versions of Firefox / Chrome / IE.

This is a low priority, having it work well on one browser is more important for the demo.

Task 1.2 - Create 'pulse' loop that will periodically collect metadata from one of the libraries

Improve server side pos analysis to try and reduce wobble

Currently the position can 'wobble' between 2 adjacent regions (and rarely across the entire room). Add in some constraints to minimize this inaccuracy. In essence create a 'window' that will vote on position based on previous measures. Also create a speed limiter that will keep people from moving long distances in short periods of time. Ie. people are limited to movement of 1 m/s and if they appear to move faster than that cap their position such that the speed constraint is not violated.

Task 8.1 - Setup sever on web space

Task 9.4 - Design and implement 'About' page

Task 9.1 - Setup web framework on server

Task 2.5 - Put some consideration into limiting metadata transfer to take into account possible mobile bandwidth limits

Task 3.1 - Find and configure compatible hardware. Either the Bluetooth Dongles in the Software lab or 3rd party hardware

Task 9.7 - Implement 'Analytics' page

Task 9.2 - Implement authentication on web site

Task 9.5 - Implement 'Live Feed' Page

Depends on (4.5)

Task 1.3 - Receive the metadata from the libraries, bundle it up and send to Tier 2 REST Server

Task 5.4 - Push notifications for customer events sent to employee app

Task 2.4 - Send metadata to REST position server. We will have to play with timing and frequency to see what works best for the desired results

Task 9.6 - Implement 'Time Lapse' page

Depends upon (4.5)

Task 6.3 - Filter metadata collected to only include callibration device

Depends on (4.5) and (4.3)

Task 7.3 - Display data pull from customer tracking system on 'live display'

Switch Android App to use a Bluetooth API that is available on 4.2.2

Currently the Bluetooth data collection requires Android 4.3 (API level 18). Lab phones only support API 17 (Android 4.2.2). We need a BT collection solution that works on those phones. Easiest approach should be to update our code. Plan B is somehow convince Dr. Dick to let us upgrade the phones software. Nexus 4 can use at least 4.4.

Task 6.5 - Send heat maps to Position System

Task 1.4 - Soak test drone

We may have some issues with running the drones for long periods of time. Once we have more of the system in place I want to do a soak test and see if we hit any issues. Airodump-ng can apparently have problems with certain processes that we may not be able to kill. See Task 1.1 comments for details.

Fix position system FSPL equation to use variable frequency

FSPL equation is dependent upon the frequency of the signal. However, currently a fixed frequency is assumed as WiFi is always the same. To support shared code between WiFi and BT position solving the calculations need to account for the frequency of Bluetooth signals if Bluetooth signals are the source of the data.

Task 7.4 - Receive notifications regarding customers that haven't been helped recently

Task 6.4 - Build heat maps for each sensor

Task 4.3 - Store raw metadata in database

Depends upon (8.2)

Task 9.8 - Implement 'Settings' page

Task 2.2 - Implement signal strength metadata collection to Bluetooth device form all beacons within range.

Task 5.3 - Data to store customer tagging

Depends upon (8.2)

Task 1.1 - Download and evaluate Air Crack and Snoopy libraries

Don't make the TA's life hell

Setup is a bit of a bitch at this point. We need a few things so the TA can do some testing of our code. Firstly setup scripts need to be created wherever possible, this should work for server configs and some of the WiFi stuff. We also need write ups on setup and use of all parts of our system. Finally a ready to go system image would be great to have. The goal is to make the TA not hate us when he sees just how much stuff we used to build our system.

Task 5.5 - Improve position calculations using calibrated data

Task 6.1 - Study mathematics of variable signal strength