A comparison of file-system and drone fleet software levels.

A software centric approach is required for the purpose of relaying commands and treating the drone formation as a single entity.  A multi-layer approach is helpful to articulate the logical high-level commands to actual unit movements. This is the same approach used for file systems where disk access is almost hidden from the user level file access. The management and access must be separated. It is helpful to organize this in multiple layers where the lower layers can be switched out with those applicable to specify types of drone units with restricted particle movements while the higher-level layers remain in place. The higher-level layers also describe manageability capabilities for drone unit and group such that they can be accessed interactively or programmatically. The following article compares the file-system organization levels with the drone fleet management software levels.

File-System organizational level	Drone Fleet organizational levels
Application Programs -          These are the consumers of the file-system organization and access all its interfaces such as for create, update, delete of directories and files	Application Programs -          These are the consumers of the fleet management organization and use all its interfaces such as for co-ordination of group and individual drones.
Logical File system -          This manages metadata information about a file that helps with its bookkeeping as well as the file control block which records owner, size, permissions and locations	Logical Fleet system -          This manages metadata information about a drone that helps with its bookkeeping as well as the capabilities of a drone such as in terms of its flexibility to perform 360-degree movement.
File Organizational module -           This has information about the files, the location of the files and the logical and physical blocks. While logical blocks could be contiguous, the physical blocks may be scattered with free space between them.	Fleet organizational module -          This has information about the drones, the hierarchical addressability of the drone and the logical and physical formations. The formations can vary independent of the fleet so that the same fleet can have different formations. One formation can transform to another with group commands.
Basic File system -          This comprises of the general commands to the device drivers to read and write physical blocks on the disk. Blocks provide a convenient abstraction from user organizational primitives such as folder and files as well as a system defined data structure that can record information about the physical layout on disks	Basic fleet system -          This comprises of the general command to state reconciliation and operations on the fleet which translate to physical commands to drones. Separation of drone commands in terms of system defined data structures and methods is beneficial to allowing a variety of drones as the individual units of the fleet even though the fleet may be homogenous.
I/O control -          Device drivers act as an interface between the devices and the higher-level layers. With a block number and low-level hardware specific instruction, device-drivers facilitate compliance to a common programming interface.	I/O control -          Drivers dedicated to the drone for its make and model such that it behaves consistently with the expectations of higher-level layers allow them to be switched one for another.
Devices -          The ultimate store for the bits and bytes of the user-defined files and folders are the disks and they could comprise their own stack so long as the devices behave as a permanent store	Drones -          The ultimate low-level for a fleet management comprises just the drones that can come with software stack dedicated to handling common commands for its movements.

 

The scheduling and control are handled at the fleet level as well as the drone level by modules at the corresponding levels. The generic displacement commands to a drone can be articulated in a sequence of motions to achieve that displacement by the drone level scheduling and control unit when it has multiple moving parts – each of which must be controlled with the precise positioning and orientation.

While autonomous decisions are made by individual drones, they must remain within the overall trajectory tolerance space for the entire fleet. This can be achieved with the popular neural net for softmax classification. The goodness of fit for a cluster or sum of squares of errors of F-score are other alternative measures.

#code

using System.Collections.Generic;

 

namespace SquadronFormation.Abstractions

{

    /// <summary>

    /// Represents a Formation

    /// </summary>

    public abstract class FormationBase : DroneFormationBase

    {

        /// <summary>

        /// Enumerates all units in the formation.

        /// </summary>

        /// <returns>Collection of Drones and Formations</returns>

        public abstract IEnumerable<DroneFormationBase> EnumerateDroneFormations();

 

        /// <summary>

        /// Returns an instance of <see cref="FormationBase" /> that represents a formation within the overall

        /// </summary>

        /// <param name="path">The Formation reference</param>

        /// <returns>Instance of <see cref="FormationBase" /> even if formation does not exist</returns>

        public abstract FormationBase? GetFormation(string path);

 

        /// <summary>

        /// Returns an instance of <see cref="DroneBase" /> that represents an atomic drone in the formation

        /// </summary>

        /// <param name="path">The drone qualifier</param>

        /// <returns>Instance of <see cref="DroneBase" /> even if drone does not exist</returns>

        public abstract DroneBase? GetDrone(string path);

    }

}