Exercise 2: Needs Assessment

• Problem: Before going to a campus facilities like they gym, it would be nice to know how crowded it is ahead of time and if it is, when would be a better time to go. This would allow you to make better use of your time and better plan your day.
• Objective: Being able to check a website with current traffic estimation at a facility and when peak and low hours are based on historical data.
• Background: This project will be a joint CPE and EE effort as it involves programming in conjunction with hardware to create a usable product.
• Methodology: To complete this task the raspberry pi would be stationed in the area which the traffic is to be estimated. The pi would host a Django web server that users could check the current and historical traffic. To estimate the traffic the pi would be equipped with a Bluetooth dongle. The dongle would scan the proximity for Bluetooth signals coming from peoples phones to gauge how crowded the area is.
• Expected Results: Website with current traffic estimation with qualitative measure (busy, moderately busy, slightly buys, not busy). The website will also show historical averages of hourly traffic on each day and show basic information about the facility like hours of operation.
• Costs: Raspberry Pi, Bluetooth Dongle, Case
  • ~$45 total

Exercise 3: Problem Formulation

Exercise 4: Solution Development

• Versatility: The device should be easily configured to be deployed to any location. The main variable will be the size of the location. For smaller locations, only Bluetooth signals with strong signals should be accounted for, larger locations should include weaker proximity and for very large areas (i.e. the library)the device should be able to coordinate with multiple auxiliary scanners to get an accurate measure of traffic.
• Reliability: In the event of power or internet loss the device should be backed up and automatically resume when it is able to come back up.
• Cost: The device should be low cost so it can be widely deployed to give users information about all public facilities.
• Customization: The web interface should be easily customized to fit the needs of a specific facility in appearance and available data.
• Data Analysis: Be able to interpret and draw conclusion to make the raw data useful. The data analysis should be customizable as well for specific locations (i.e. library traffic will be higher during midterms and finals)

Exercise 5: Intellectual Property

Exercise 6: Abstraction and Modeling

Exercise 7: Synthesis

Exercise 8: Ethics and Liability

Foreseeable Misuses

1) Sensors used could be faulty and update the servers when nothing should be inputted

2) Hacking to change the current numbers for each facility

Possible Changes

1) Students can have an input on the app itself

2) The feed could become self-updated by the staff in charge of the facility at the given time instead of sensors

Disposal after the useful life has ended

1) Taking the data taken from all of the years in its useful life to comprehend the day-to-day life of college students

2) All the technology used must be properly disposed of

Exercise 9: Failure and Hazard Analysis

Solutions to potential hazards proposed in Exercise 8

• Create a firewall or implement a form of cyber security to protect the servers from potential hackers
• Have the staff in charge of the facility monitor the live-feed every hour to make sure the numbers are correct and allow them to correct if an error were to occur
• Allow students to have an input on the app to allow for useful discussions about facility usage

Exercise 10: Design Analysis

Following the morphological chart, the team began to analyze various design alternatives, specifically the radio receiver that will be used to detect different cell phone and bluetooth device signals that will be used to predict how crowded an area is. The team narrowed down the alternatives to SDR options and compared three options within SDR-type using a decision matrix shown above. The analysis showed the low-end option was superior to others because of its mainly because of its low-cost and small form factor advantages. An RTL2832U based dongles retail anywhere from $5~$20 whereas the Ubertooth One sells for $125~$150 and hackRF for aroud $300. Additionaly the RTL2832U dongles have an form factor advantage compared to the other two options, which in returns helps the portability and weight of the final product.