This project is aimed to provide a customized solution of quicker and more satisfying food order and delivery system under the digital business trend. For the food order part, a “Pizza Hot” website and mobile applications are implemented for customers to view dishes and place orders. Meanwhile, restaurant managers are able to manage dishes, orders, comments by logging in the website management dashboard. For the food delivery part, an intelligent delivering system is built, including a mobile application for delivery staff and a smart delivery box integrated with a micro-controller temperature sensor, WiFi and GPS modules, which can show temperature of pizza, time, and current location on the screen as well as navigate delivery staff to next delivery by audio output.
This MEng project is to implement and extend a business model proposed by four students in Digital Businesses Strategy class from Johnson School. It is an events finder targeting specifically on all the events in Cornell and college town. It contains a website, an iOS application and a ticker. The skill needed for the project are web full stack development, iOS application development, cross platform database management and embedded system development. A website and an iOS application which support events displaying, searching, posting and cross platform data synchronization has launched. A Raspberry Pi based arcade machine which can show upcoming events and allow people to play game with the joystick and buttons on it is also built. Compared with similar events finders, this event finder could overtake them because it is more informative, highly localized and cross-platform.
Dongle, Database, and GUI for Managing Unique Bird Tag IDs
Our project is an interdisciplinary project including Biology and ECE fields. Our team has four members and we are responsible for the ECE part. First of all, we developed a USB dongle containing a receiver to detect tag codes and a GPS receiver to record location. As for the tag part, we designed a new tag with a smaller solar beeper which can utilize the solar energy more efficiently. In addition, we also extended and refined the GUI and its interaction with the cloud- based database. After the project is finished, field biologists can use this device with a laptop computer to capture tag codes, times and locations as they are deployed, along with species and band number of the animals being tagged. Whenever a tag ID is recovered through RF communication in the field, similar data will be recorded and stored to the cloud-based database. It is a very meaningful project for protecting the endangered birds. What’s more, this project can also be applied to other endangered animals.
Real-Time ROSberry Pi SLAM Robot
A realtime monocular (single camera) visual Simultaneous Localization And Mapping (SLAM) robot is built utilizing a server-node based computational approach. The core of the robot is a Raspberry Pi 2 with a Robot Operating System (ROS) wrapper over the Raspbian Wheezy Linux kernel. Different nodes from the robot communicate with the server to map its location based on its surroundings using ORB-SLAM, with loop detection and re-localization capabilities.
Foot Motion Capture
At the present time, there is only anecdotal evidence concerning possible design issues with Firefighter boots. This motion capture system is used to define the possible problems when Firefighter boots are being used. The system has three functions: data detection, data collection, data display. SV03A rotary sensors are being used in this system to detect the angle change data of ankle of the foot. Its biggest advantage is that it fits into the small room of Firefighter boots. Data collection part uses Arduino Uno micro-controller to realize its function. Once the angle change has been detected by the sensor, micro-controller will filter the data and transfer it from voltage to angle data and finally store it locally on a SD card by using the SD card shield. Last function for this system is to display data in animation form. After finishing collecting data in an experiment, data will be uploaded to a Processing program and display the data out in an animation form.
Parallel Programming in a Raspberry Pi Cluster
The aim of the project is to build a Symmetric Multi-Processing (SMP) and Asymmetric Multi- Processing (ASMP) Platforms to develop Parallel Applications. The Project is planned to be used as a Lab project for the course “ECE5725 – Design with Embedded Operating System”. Aimed at providing the students with exposure to SMP and ASMP in an inexpensive but solid way, the Project was built using Raspberry Pi 2 Model B embedded computers. Raspberry Pi 2 being a four core processor with shared physical memory acted as an excellent platform for SMP. Meanwhile, for ASMP, a server cluster was built by connecting four Raspberry Pi’s using an Ethernet network router. Sobel Filter Edge Detection was chosen as a target application to analyze the performance and also to be used as a lab exercise for the course. Sobel filter with its two Dimensional array computation turned out to be an excellent application to learn the corner cases that are associated with the parallelization of a program. SMP and ASMP Programs, both individually and also as a hybrid, achieved the four and ten times better performance than sequential programming respectively.