Authors:
Manda Miller, Albert Bennah, Ashley Wehr, Kristen Chung, Akshay Pandit
Abstract:
A unique method is proposed to ensure accurate delivery of eCommerce orders that uses a combination of longitude/latitude information from metadata, image characteristic (ex. identifying objects and other features in the image), and nearby image-producing IoT home devices to alert a delivery driver when they are at incorrect location and prevent incorrect delivery of a package.
Background:
When purchasing online or via ecommerce where you get a delivery to your home, sometimes the order and/or items are delivered to a neighbor rather than the correct house. Delivery services have implemented a process where they take a picture of the door of the house, or porch, or other location at your home to show you that the item was delivered. The picture validating the delivery is not helpful if the picture is of someone else’s door or porch. The result is frustration for the package receiver who must now try to locate their incorrectly delivered package or file a complaint with the shipping company.
Description:
The delivery driver pulls up to a home that he/she has determined is the place to deliver the package.
2. Delivery driver has the package, they scan it, and proceed to leave it on the porch
a. The scan location or address is logged.
3. After leaving the package on the porch they take out their company designated mobile device to take a picture of the proof of delivery
a. Picture is taken, and longitude/latitude is logged and cross referenced to the longitude/latitude of the address (Google Maps even lists a long/latitude for an address. See below the longitude/latitude listed for our Toshiba Global Durham HQ office.
35.90738 x -78.84490
b. Picture is then cross-referenced to a database of other pictures taken for that address, and the corresponding longitude/latitude logs within the metadata.
c. If an IOT device is present, such as a Ring or Nest doorbell, this device can be attributed, via a user profile for the delivery service, to that person’s address. The IoT device can then send info of the longitude/latitude of the metadata as a reference to use when comparing longitude/latitude of the delivery driver image.
4. An algorithm compares the data and determines, within a certain degree of certainty, that the location matches all the other times a delivery was logged for this location (pictures are similar), and that within a certain degree of certainty, the long and latitude is correct for the address.
5. If all is deemed a match, the delivery driver can close out and complete the delivery.
6. If all is not deemed a match, the delivery driver mobile device activates a notification and message to the driver that it is the incorrect location. The driver cannot mark delivery as complete, and the images are not accepted until the driver corrects the delivery location.
7. The system then provides a map to the correct longitude/latitude of the address and redirects the driver to a more accurate location or either the delivery driver or automated system contacts the customer to inform them of a designated location to pick up the package such as a locker or delivery service facility.
TGCS Reference 00176