top of page
00.png

Movable Delivery Pick-up Station

​As time goes on with advanced technology, drones are integrated into our society and can be deployed in the delivery industry with autonomous flying. This futuristic drone delivery simulation project is based on an agent-based model framework. Each drone in the system represents an agent to execute the pick-up task without human control under the definition of the behaviors.

Buildings

01.png

Boundary Condition

Drone

​Delivered Station

Mothership

Mothership

Drone

​Delivered Station

This system has three agents: Mothership, Delivery Drones, and ​​Delivered Stations. The Mothership is the main container for the drones, functioning for launching the drones and charging. The Drones are released from the Mothership and find the destinations to reach. They will keep searching for the destination nearby and try to hit as much as possible until the battery capacity is lower than the threshold. When the battery capacity is lower than the threshold value, it will return to the Mothership immediately for charging.

DataDiagram_4x.png
System Structure_4x.png

Here is the overall framework system; it is structured with three types of agents. Each of them has different properties that will affect each other.

Topview.gif

The Mothership has the highest hierarchy in the system, determines when the drones will be released, and will move toward the mean of either drones or destinations to shorten the length of the total path distance potentially. By counting the current operating drones, the "Following Movement" function will lead the Mothership toward the means of drones or destinations. The" Restriction" function is to constrain its boundary condition and also restrict the flying height of the Mothership itself. The "Release Drones" function determines the condition when the drones should be released; currently, it's based on the iteration count, but it could potentially add on other conditions to make the releasing movement more dynamic.

Releasing.gif

For the Drones behaviors, it contains five functions, "Determine Destination," "Flying Mode," "Flying Execution," "Avoiding Buildings," and "Color Display."
The "Determine Destination" function initially calculates the closest destination for the drone to target.
For the "Flying Mode," the system has a checklist of toggles, such as operating status and battery data, to determine the current moving status of the drone itself. And the drones will change their flying movement according to their status. The charging status toggle is based on a battery counter that will trigger the toggle when the counter exceeds or is reset. The operating status will turn true when it is released and turn to false when it returns to the mothership. Only when both operating statuses and the full battery status are true will the drone start changing the flying mode and switch between searching, landing, and lifting. So after we have the toggles of all flying modes, the drone movements are executed by assigning the direction to the velocity.
Third, "Avoiding Buildings" is an avoid buildings function for crashing and collisions.
Lastly, in "Color Display," the drones will visualize different color schemes depending on the flying modes. When the drone hits the destination, it turns green and keeps searching and flying to the next closet spot while it still has enough battery. After reaching the second destination, the battery capacity is insufficient, and the color will turn red while going back to the mothership for charging.

The last agent is the destinations. This is for re-generating the same number of destinations in the system for drones to target.
The destinations are displayed as glowing yellow rings. When it is reached by drone, the yellow light vanishes, and the other destination light on chose randomly.

Simulation Start from Drones Number: 10

Simulation Start from Drones Number: 20

Simulating the drone behavior leads us to output the chart showing the relationship of different parameters by time increasing, which can potentially be used for further optimization.

For the future potential next step, it could increase into multi types of drones with different speeds and battery capacities and flying zone conditions, which adds up to another layer of complexity to this drone system.

Simulation Start from Drones Number: 30

Original Simulation Animation

bottom of page