VVFGeoFleet is a system allowing to track a fleet, where vehicles are equipped with GPS enabled devices periodically sending their position.

The system is composed of two main modules:

• a backend module, based on a RESTful server;
• a frontend module, based on an Angular GUI.

The backend is written in C# Asp.Net WebApi, using Visual Studio 2017 Community Edition, with a persistence layer based on MongoDB document database.

The backend:

• collects position messages coming from GPS enabled devices installed on board the vehicles;
• publishes REST actions to efficiently query the collected data.

Queries allow, for instance, to:

• getting the current fleet position (also by vehicle class);
• getting a vehicle position;
• getting a vehicle tracked path, within a given time interval;
• getting vehicles near to a given point (also by vehicle class);
• getting vehicles within a given rectangle;
• getting vehicles inactive since too much time.

The backend is optimized to collect data and hold them forever, without the need for purging old data periodically. The RESTful API leverages the MongoDB database and its spatial features. All database queries are carefully optimized to exploit indexes; thus, the response time is extremely low even in case of millions messages stored.

The RESTful API enables the frontend to graphically display data, but it is especially conceived to be queried by other applications (B2B) in order to act as a decision support system (DSS).

The frontend is based on Angular4 and uses Google maps javascript API to show information.

To be completed

Currently the source code and the RESTful interface is a mix of English and Italian words. During the implementation, the Italian words will be translated as much as possible.

VVFGeoFleet can receive keep-alive messages (e.g. every minute or so) from vehicles, in order to promptly know whether a vehicle is out-of-field or the GPS device is faulty. In presence of many vehicles (i.e. more than 10.000) those messages might be heavy to be stored.

To cope with this issue, VVFGeoFleet holds only messages carrying position deltas, while interpolating messages received by the same vehicle in the same position. Interpolation information is saved together with the interpolating messages, so to prevent such inforation to get lost.

On each position message reception, VVFGeoFleet computes the vehicle velocity comparing the received message with the latest stored one. In case such velocity exceeds a given threshold, a warning is traced in the application log (based on log4net). Too high velocities might be due to bad GPS device configuration of faulty GPS device.

System publishes statistics about message arrival per day, total number of stored messages, messages arrival rate, etc. Statistics can be read to check correct system operation, and can be processed through automated systems in order to raise alarms in case of anomalous indicators.

Posting position messages can be subject to client authorization, based on IP source address. The web.config allows to enable/disable IP-based authorization through a flag and to specify the list of allowed IP and/or networks (e.g. 127.0.0.1, 8.8.8.8, 10.0.0.0/24, 172.16.0.0/16).

The server exposes a web socket the clients can subscribe to, in order to receive new vehicle positions, as they arrive, via push notifications.

Stores a new position message. Position message payloads are structured as follows.

{
    "codiceMezzo": "APS.12345",
    "classiMezzo": [ "APS", "benzina" ],
    "localizzazione": {
        "lat": 14.2345,
        "lon": 43.4321
    },
    "istanteAcquisizione": "2017-11-27T10:12:00.1234Z",
    "fonte": {
        "codiceFonte": "AH222",
        "classeFonte": "gpsTracker"
    },
    "infoFonte": {
    
    }
}

It returns the stored message location, together with the message itself.

Returns the message with the specified id.

Returns current position for the specified vehicle code (translated as codiceMezzo in the code).

Note: should vehicle code contain a dot (.) character, you have to add a trailing slash (/) to the action call. Otherwise you would end up having a 404 error.

Returns current position of the entire fleet.

Only the updated positions are returned, i.e. those arrived since the last attSec seconds. attSec parameter is optional and defaults to the value contained in the Web.Config file (orizzonteTemporale_sec parameter).

Returns current position of the entire fleet, just for the vehicle classes specified as input parameters.

Only the updated positions are returned, i.e. those arrived since the last attSec seconds. attSec parameter is optional and defaults to the value contained in the Web.Config file.

Returns vehicles close to the specified point, within the maximum radius specified as input parameter (distanzaMaxMt).

Optionally, it is possible to specify an array of vehicle classes as a filter.

Only the updated positions are returned, i.e. those arrived since the last attSec seconds. attSec parameter is optional and defaults to the value contained in the Web.Config file. The distance from the specified center is returned, too.

The result is structured as follows.

{
	"istanteQuery": "2018-03-04T20:29:03.686917Z",
	"punto": {
		"type": "Point",
		"lat": -44.630001068115234,
		"lon": 3.4300000667572021
	},
	"distanzaMaxMt": 300000.0,
	"numeroMezzi": 2,
	"durataQuery_msec": 5,
	"risultati": [{
			"messaggioPosizione": {
				"id": "5a9c24f74eb6d93300a3e2a0",
				"codiceMezzo": "M.1101",
				"classiMezzo": ["AV", "diesel", "nero", "metallizzato"],
				"localizzazione": {
					"type": "Point",
					"lat": -44.7388,
					"lon": 3.2939
				},
				"istanteAcquisizione": "2018-03-04T01:23:46.336Z",
				"fonte": {
					"codiceFonte": "J55",
					"classeFonte": "C2"
				},
				"infoFonte": null,
				"infoSO115": {
					"stato": "InViaggio",
					"codiceIntervento": "11KXS",
					"dataIntervento": "2017-10-24T11:33:16.974Z"
				},
				"istanteArchiviazione": "2018-03-04T16:55:19.424Z",
				"ultimo": true
			},
			"distanzaMt": 16208.5889
		}, {
			"messaggioPosizione": {
				"id": "5a9c24ff4eb6d93300a3e822",
				"codiceMezzo": "M.1236",
				"classiMezzo": ["AV", "diesel", "amaranto"],
				"localizzazione": {
					"type": "Point",
					"lat": -47.0248,
					"lon": 1.6938
				},
				"istanteAcquisizione": "2018-03-04T14:12:18.937Z",
				"fonte": {
					"codiceFonte": "A09",
					"classeFonte": "C1"
				},
				"infoFonte": null,
				"infoSO115": {
					"stato": "SulPosto",
					"codiceIntervento": "79WDW",
					"dataIntervento": "2017-07-27T21:44:36.896Z"
				},
				"istanteArchiviazione": "2018-03-04T16:55:27.099Z",
				"ultimo": true
			},
			"distanzaMt": 298653.844
		}
	]
}

Returns vehicles within the specified rectangle.

Optionally, it is possible to specify an array of vehicle classes as a filter.

Only the updated positions are returned, i.e. those arrived since the last attSec seconds. attSec parameter is optional and defaults to the value contained in the Web.Config file.

Returns position messages for vehicles not updating their position since too long time. Useful to detect anomalies on GPS devices.

For instance, if the daSecondi is equal to 86400, the action returns the last position message for each vehicle not sending its position since at least one day.

Returns position messages for vehicles not updating their position since too long time. Results are filtered by vehicle classes specified as input parameters.

Returns all vehicle classes with their occurrence for all active vehicles (i.e. those which sent their position within the specified seconds).

This is a faceted search useful to have information about vehicle classes, especially useful to enable fleet filtering by class.

Results are returned in the following form.

[
  {
    "class": "AV",
    "count": 517
  },
  {
    "class": "APS",
    "count": 322
  }
  //...
]

Returns statistics about system operations (i.e. numer of messages processed per day, total number of message stored, message arrival rate, etc.). Statistics JSON format is the following:

{
	"lastNews": {
		"messages": {
			"numberOfMessagesStoredInTheLastMinute": {
				"net": 100,
				"withInterpolation": 1000
			},
			"numberOfMessagesStoredInTheLastHour": {
				"net": 6000,
				"withInterpolation": 60000
			},
			"totalNumberOfMessagesStored": {
				"net": 1000000,
				"withInterpolation": 10000000
			},
			"averageNumberOfMessagesPerMinute": 100.0,
			"averageNumberOfMessagesPerSecond": 1.6667
		},
		"vehicles": {
			"numberOfVehicles": 1000,
			"numberOfVehiclesActiveWithin72h": 1500,
			"numberOfVehiclesActiveWithin48h": 1400,
			"numberOfVehiclesActiveWithin24h": 1300,
			"numberOfVehiclesActiveWithin1h": 1200,
			"numberOfVehiclesActiveWithin1m": 50
		}
	},
	"dailyStats": [{
			"key": {
				"year": 2018,
				"month": 3,
				"day": 20
			},
			"net": 10000,
			"withInterpolation": 100000
		}, {
			"key": {
				"year": 2018,
				"month": 3,
				"day": 19
			},
			"net": 12000,
			"withInterpolation": 120000
		},
		...
	]
}

Returns the path tracked for vehicle having the code specified as parameter, within the specified time interval.

Date are represented in ISO-8601 format (i.e. 2018-03-04T10:45:52.875Z).

Forwards every message received from /api/messaggiPosizione through a WebSocket. WebSocket is implemented through the SignalR library.

VVFGeoFleet backend depends (also) on the following libraries.

• MongoDB.Driver: C# driver for MongoDB;
• SimpleInjector: Dependency Injection (DI) library;
• log4net: application log;
• NUnit v3: unit tests library;
• Moq: mock classes generation library;
• Bogus: fake data generation library.
• Microsoft.AspNet.SignalR: Incredibly simple real-time web for .NET. This package pulls the server components and the JavaScript client required to use the SignalR in an ASP.NET application.
• Microsoft.Owin: Provides a set of helper types and abstractions for simplifying the creation of OWIN components.

VVFGeoFleet frontend depends (also) on the following libraries.

• @agm/core: Angular Google Maps;
• bootstrap: Bootstrap css library;
• moment.js: date and time handling routines;
• typescript: dev library enhancing javascript to be OOP.

Source code is released under the terms of AGPL-3.0 license.

Use this project at your own risk. The authors are not responsible for any damage which might result from this project usage.