Proud to Have Reached MIT

The 2017 spring course in “Interconnected Embedded Systems” at MIT required the submission of a group project. One of the many creative student-efforts that were submitted has been presented in detail on freeCodeCamp by Moin Nadeem, one of the creators of the project. Along with Avery Lamp, Daniel Gonzalez and Ethan Weber, Nadeem built a device which monitored Wi-Fi probe requests, that is, snippets of such information as a unique MAC address, RSSI signal and a list of previous SSIDs encountered. By leveraging this information, students around the MIT campus could be tracked, in order to analyze trends, such as most frequently-visited places on the MIT campus, peak times of student traffic, etc. However, all data collected were anonymized. These data were not fine-grained enough to determine a mapping from MAC addresses to individuals, thus mitigating any privacy concerns one could have.

The students on the project, then, focused on utilizing FTDI programmers in order to flash an implementation of Arduino for ESP8266. This enabled them to continue with an environment which provided a breadth of libraries of the built-in AT-command firmware.

This was then complemented by an infrastructure which would permit collecting probe requests en-masse. A Flask + MySQL backend managed the device infrastructure and information, an iOS application eased the deployment of devices, a front-end to the relevant website was created, as well as an analytics platform which transformed the wealth of incoming data into intelligible data with valuable insights. On the hardware front, ESP8266 chips were soldered onto prototype bards, along with some power modules. The resulting devices were made entirely portable with the use of PowerBoost 1000Cs, which also permitted the device users to perform tracking in some less accessible locations.

After a code for auto-connecting the devices to the nearest unsecured Wi-Fi hotspot on boot was written, an application was also constructed for updating the device location. This was useful for knowing which MAC addresses to associate with each location.

Analyzing the received data showed a periodic student behavior behind each location they visited on campus. Furthermore, some greater trends throughout the campus were highlighted. For example, major arteries achieved peak traffic at around 5 o’clock in the afternoon, whereas buildings on the edge of campus achieved peak traffic at noon. In fact, the data-science possibilities in this project are endless. This technology could be leveraged to create smarter cities, work against congestion, and provide better insights into how people could be able to reduce mean walk times. 

Congratulations to all the students who built this project at MIT!

Have you created any exciting projects, using Espressif products? If yes, let us know, so we can share them with the whole wide world!