Project
| # | Title | Team Members | TA | Documents | Sponsor |
|---|---|---|---|---|---|
| 22 | Smart Drone Delivery Improvement |
Rahul Joshi Raymond Hoagland Sachin Weerasooriya |
Luke Wendt | appendix0.docx design_document0.pdf final_paper0.pdf other0.pdf presentation0.pptx proposal0.pdf |
|
| When you order a package that is shipped via ground, it is either left at your door or you must be present to sign off on the package. Services like Amazon PrimeAir look to speed up delivery time with drone shipping, which claims to be able to come to your door in 30 minutes. If you look at Video 2 in this link, you will see that the shipment is loaded and the drone takes off, converts to a plane, flies to the vicinity of the landing local, converts back to a drone, and lands on a marker put out by the recipient. This drone then deposits the package and returns to the factory for its next package. Here lies a major flaw; in the event that something valuable is being shipped, it would be desirable for their to be a confirmation that someone is available to pickup the package that is being delivered. This is where we step in. The problem we want to address is we will assume that the drone is in the area of the delivery spot. We will use image processing to ID the user specified landing spot. The key difference will be: instead of landing, the drone will notify the owner that the package is ready for pickup and hover above the landing spot for a fixed amount of time. The drone will then wait for a confirmation from the user that it is safe to drop off the package. If this message isn't received after a set amount of time, the drone will return to the warehouse with the package and will try to return the package later. While the drone waits, it will constantly scan the surroundings to see if any unidentified threats are approaching the drone. If it detects a threat is too close, the drone will take off and hover at a higher elevation to protect the package contents. It will stay there until either the recipient gives the okay for delivery or until the time limit is up and the drone will return to the factory. Hardware needed: - A drone - A small camera - Digital Signal Processor (DSP) Chip - Proportional-Integral-Derivative (PID) Controller - Raspberry Pi - USB wifi dongle - Smart phone We will use image processing to ID our landing spot and the PID controller to send the necessary feedback to the motors in order to descend the drone and lookout for potential hazards. For our prototype, will then use the raspberry pi and dongle to connect to the resident's WiFi and send a message to a smartphone app indicating the package has arrived. The drone will then wait for confirmation from the app. The mechanical mechanism to physically lower the package will be out of our scope. Rather we will have an LED or some way to indicate that the package has been dropped. |
|||||