AI Smart Transportation

The design of a Smart Transportation system requires the use of reliable plugins capable of operating in the most challenging contexts, typically outdoors, in unpredictable weather and lighting conditions.

To this end, A.I. Tech provides you with a wide selection of plugins that, when combined in the right way, can provide a high-quality response to your design needs. They are designed using the most modern Artificial Intelligence methodologies and developed by optimising the code in such a way that, while providing the same functionality and image resolution, they are extremely fast.

One of their special features is that they can be used on multiple hardware platforms: edge side (on cameras), server side, or on the AI Appliance platform. Their integrability is practically unlimited, thanks to the support of most of the communication protocols used today, such as Modbus/TCP and HTTP, to drive Remote/IO devices, as well as about a dozen third-party VMSs.

AI-Traffic

AI-TRAFFIC is a video analytics application based on the most advanced artificial intelligence algorithms, that includes all the functionalities of AI-ROAD3D and AI-INCIDENT, thus meeting the needs in the smart road field.
AI-ROAD3D counts and classifies vehicles crossing virtual sensors in a given direction. Three classes of vehicles can be identified: motorbikes, cars and trucks. The analytics is also able to estimate the color and average speed of each vehicle and fires an alarm if this speed exceeds a customizable threshold. It also allows real-time estimation of traffic density as well as the monitoring of traffic flows through origin-destination matrices.
AI-INCIDENT detects anomalous and dangerous situations on the road, such as vehicles driving on the wrong side of the road, stationary vehicles, U-turs, lane changes, queues and pedestrians in forbidden zones.
AI-TRAFFIC combines an advanced 3D calibration and reconstruction mechanism of the scene with the most advanced artificial vision and artificial intelligence algorithms.
Like the AI-ROAD3D and AI-INCIDENT apps, AI-TRAFFIC uses the latest deep learning algorithms for both object detection and classification (distinguishing vehicles and people) achieving a high accuracy even in extremely complex scenarios, such as in tunnels or crowded city streets, at night or in severe weather conditions.

AI-TRAFFIC(-DEEP) include le funzionalità delle differenti app pensate per rispondere alle più svariate esigenze di ogni città che voglia definirsi “smart”. L’app infatti consente di comprendere e analizzare i flussi veicolari nelle varie arterie cittadine, attraverso il conteggio delle varie categorie di veicoli. Analizzare la velocità media dei veicoli sulle varie tratte inoltre consente di definire le strade attraversate con una velocità media più elevata (eventualmente superiore rispetto ad una soglia impostata), suggerendo così una posizione ottimale per il posizionamento di pattuglie o di sistemi automatici che possano essere utilizzati per fini sanzionatori. L’app consente inoltre di identificare situazioni di potenziale pericolo su strada, come ad esempio un accodamento o un veicolo che attraversa una strada contromano, o ancora la presenza di un pedone su strada.

AI-TRAFFIC(-DEEP) può anche essere utilizzato per il monitoraggio dei tunnel o in ambito autostradale.

AI-TRAFFIC is a multi-platform application, and can be installed on more than 20 different hardware platforms including cameras (edge), small processing systems (embedded devices), workstations and servers.

Edge

Embedded

Server

If you have chosen to use a workstation or server-based solution, please connect to our system design support portal: AI-DESIGNER.

AI-Traffic supports all data and alarms of AI-Incident and AI-Road3D applications

AI-Traffic supports over 20 notification protocols, including third-party VMS, standard protocols (such as HTTP, FTP, MODBUS and MQTT) and A.I. Tech proprietary protocols, which allow the notification of events to the A.I. Tech dashboards. You can get more details on the notification mechanisms available at the following link.

AI-Traffic allows you to send alarms to over 10 VMS, where you do not have a VMS you can use AI-DASH-PRO thanks to the integrated alarm management functionality.

Installation constraints

Video tutorial

Online help

Flyer

AI-Violation

AI-VIOLATION is a video analytics app that makes it possible to detect traffic light violations, i.e. vehicles that cross the stop line when the light is red.

The application also allows the identification of the vehicle that has committed this infraction, its vehicle type among the categories of car, motor vehicle and motorbike as well as average speed and the time elapsed since the red was turned on.

The detection and tracking of vehicles is based on the use of deep neural networks, as well as the analysis of the traffic light status. In fact, the application is able to determine the status of the traffic light (red, yellow, green) automatically, with only artificial intelligence applied to the processing of the video acquired by the camera, without the need for any physical connection with the traffic light.

AI-VIOLATION is a key tool for public administration, as it allows them to identify irregularities linked to vehicles passing red lights. Understanding the areas where these violations occur can be a useful indication for the public administration to decide the most suitable position in which to insert the device that will be used for sanctioning purposes.

On the other hand, AI-VIOLATION can also be considered as the ideal solution to be integrated into approved systems for fining purposes.

AI-VIOLATION is a multi-platform application, and can be installed on more than 20 different hardware platforms including cameras (edge), small processing systems (embedded devices), workstations and servers.

Edge

Embedded

Server

If you have chosen to use a workstation or server-based solution, please connect to our system design support portal: AI-DESIGNER.

AI-Violation allows you to send an alarm when a vehicle transit, at a traffic light intersection, with red light.

AI-Violation supports over 20 notification protocols, including third-party VMS, standard protocols (such as HTTP, FTP, MODBUS and MQTT) and A.I. Tech proprietary protocols, which allow the notification of events to the A.I. Tech dashboards. You can get more details on the notification mechanisms available at the following link.

AI-Violation allows you to send alarms to over 10 VMS, where you do not have a VMS you can use AI-DASH-PRO thanks to the integrated alarm management functionality.

Installation constraints

Video tutorial

Online help

Flyer

AI-LPR

AI-LPR-DEEP is a video analytics app that utilises an advanced artificial intelligence algorithm to perform licence plate detection and recognition. Thanks to the use of an innovative engine based on semantic technologies, it also enables automatic correction of license plates based on the specific nationality of the plate (*).

The solution can detect vehicles up to a maximum speed of 230 km/h (depending on the chosen hardware platform) and can be used both indoors (e.g. for monitoring car parks) and outdoors (e.g. for monitoring city streets).

* Countries for which the semantic engine is currently available: Italy.

AI-LPR-DEEP is a video analytics solution designed to meet the demands of licence plate reading. The application has various usage scenarios. The first is in car park management, as it is a fundamental tool for managing black and white lists, or even simply for associating number plates with parking tickets. AI-LPR-DEEP can also be a valid support in logistics, detecting the number plates of the various vehicles entering a port, a factory or a landfill site. At the same time, the application is also very useful in city scenarios. In fact, thanks to its ability to detect number plates at speeds of up to 230 km/h, it can be used to detect access to restricted traffic areas or access to reserved lanes.

AI-LPR is a multi-platform application, and can be installed on more than 20 different hardware platforms including cameras (edge), small processing systems (embedded devices), workstations and servers.

Edge

Embedded

Server

If you have chosen to use a workstation or server-based solution, please connect to our system design support portal: AI-DESIGNER.

AI-LPR allows license plates to be sent via over 20 communication protocols, including third-party VMS, standard protocols (such as HTTP, FTP, MODBUS and MQTT) and A.I. Tech proprietary protocols, which allow the notification of events to the A.I. Tech dashboards. You can get more details on the notification mechanisms available at the following link.

Installation constraints

Video tutorial

Online help

Flyer

AI-Rail

AI-RAIL is the video analysis application based on advanced artificial intelligence
and deep learning algorithms for the safety of railways through the detection of
obstacles on the tracks. AI-RAIL is able to distinguish the following categories of
obstacles: people, vehicles and boulders.
The application is also able to identify the presence of trains passing on the
tracks and deactivate itself dynamically; it is also possible to activate and
deactivate the application dynamically via software (API) or via hardware (via
I/O device), for example to prevent the app from generating alarms while the rail
crossing is open, and therefore allows the passage of vehicles without any
restrictions.
AI-RAIL can be used both for monitoring rail crossings, in order to verify that
they are clear, and near tunnels or rocky ridges, where the danger of falling rocks
is substantial.

AI-RAIL is the solution for making people and goods in the railway sector safe. There are two main use cases in which this application finds a natural use.
At rail crossings, it is necessary to check for obstacles (such as people and vehicles) which obstruct the passage of the train when the barrier is closed.
Therefore, it can be expected that the application is activated via external inputs, via software (API) or via hardware (via I/O device), only when the rail crossing is closed and is instead deactivated during its normal use.
Conversely, in the tunnels, when exiting them or near rocky ridges, the system must be in continuous operation. However, in order to reduce the number of false alarms, the system is expected to be deactivated during the passage of the train, thanks to the possibility of identifying the presence of a train still through artificial intelligence.

AI-Rail is a multi-platform application, and can be installed on more than 20 different hardware platforms including cameras (edge), small processing systems (embedded devices), workstations and servers.

Edge

Embedded

Server

If you have chosen to use a workstation or server-based solution, please connect to our system design support portal: AI-DESIGNER.

AI-Rail allows license plates to be sent via over 20 communication protocols, including third-party VMS, standard protocols (such as HTTP, FTP, MODBUS and MQTT) and A.I. Tech proprietary protocols, which allow the notification of events to the A.I. Tech dashboards. You can get more details on the notification mechanisms available at the following link.

Flyer

AI-Road3D

AI-ROAD3D is a video analytics application to count and classify vehicles crossing virtual sensors in real time It recognized three classes of vehicles: motorbikes, cars and trucks. The application also estimates the color and average speed of each vehicle and fires an alarm if this speed exceeds a customizable threshold. AI-ROAD3D can estimate traffic density as well as monitoring traffic flows through the origin-destination matrix.
AI-ROAD3D combines an advanced 3D calibration and reconstruction mechanism of the scene with the most advanced artificial vision and artificial intelligence algorithms.
The application uses deep neural networks to detect and classify objects achieving a high accuracy even in extremely complex scenarios, such as in tunnels or crowded city streets, at night or in severe weather conditions.

AI-ROAD3D makes it possible to meet the needs of any city that would like to be defined as ‘smart’. It provides the possibility of understanding and analysing vehicle flows in the various city arteries by counting the various categories of vehicles. Analysing the average speed of vehicles on the various routes allows the identification of roads crossed with a higher average speed (possibly higher than a set threshold), thus suggesting an optimal position for positioning surveillance patrols or automatic systems that can be used for sanctioning purposes.

AI-ROAD3D can also be used to monitor tunnels or motorways.

Finally, in combination with the AI-DASH-PRO dashboard, the app can be used to monitor car parks by counting vehicles at the gates.

AI-ROAD-3D is a multi-platform application, and can be installed on more than 20 different hardware platforms including cameras (edge), small processing systems (embedded devices), workstations and servers.

Edge

Embedded

Server

If you have chosen to use a workstation or server-based solution, please connect to our system design support portal: AI-DESIGNER.

AI-Road3D allows you to send information on vehicles that have passed through an area under analysis, in particular:

Vehicle type
Color
Average speed

It also allows you to send an alarm if the estimated average speed of a vehicle exceeds a configurable threshold.

AI-Road3D supports over 20 notification protocols, including third-party VMS, standard protocols (such as HTTP, FTP, MODBUS and MQTT) and A.I. Tech proprietary protocols, which allow the notification of events to the A.I. Tech dashboards. You can get more details on the notification mechanisms available at the following link.

AI-Road3D allows you to send alarms to over 10 VMS, where you do not have a VMS you can use AI-DASH-PRO thanks to the integrated alarm management functionality.

Installation constraints

Video tutorial

Online help

Flyer

AI-Incident

AI-INCIDENT is a video analytics application to detect in real time anomalous and dangerous situations on the road, such as vehicles driving on the wrong side of the road, stationary vehicles, u-turs, lane changes, queue or pedestrians in forbidden zones.
AI-INCIDENT combines an advanced 3D calibration and reconstruction mechanism of the scene with the most advanced artificial vision and artificial intelligence algorithms.
The application uses deep neural network to detect and classify objects achieving high accuracy even in extremely complex scenarios, such as in tunnels or crowded city streets, at night or in severe weather conditions.

AI-INCIDENT makes it possible to meet the needs of any city that would like to be defined as “smart”. It provides the possibility of identifying potentially dangerous situations on the road, such as: queuing, vehicles crossing the wrong way, or the presence of pedestrians on the road.

AI-INCIDENT can also be used to monitor tunnels or motorways.

AI-INCIDENT is a multi-platform application, and can be installed on more than 20 different hardware platforms including cameras (edge), small processing systems (embedded devices), workstations and servers.

Edge

Embedded

Server

If you have chosen to use a server-based solution, please connect to our system design support portal: AI-DESIGNER.

AI-Incident can send alerts when the following situations are detected:

Vehicles in the wrong direction
Presence of pedestrians in areas where the transit or stop of people is not allowed
Presence of stationary vehicles in the carriageway

AI-Incident supports over 20 notification protocols, including third-party VMS, standard protocols (such as HTTP, FTP, MODBUS and MQTT) and A.I. Tech proprietary protocols, which allow the notification of events to the A.I. Tech dashboards. You can get more details on the notification mechanisms available at the following link.

AI-Incident allows you to send alarms to over 10 VMS, where you do not have a VMS you can use AI-DASH-PRO thanks to the integrated alarm management functionality.

Installation constraints

Video tutorial

Online help

Flyer

AI-SmartCity

AI-SMART CITY provides the functionalities to manage a modern city, through the analysis of vehicles and people with advanced deep learning algorithms. AI-SMART CITY includes the features found in AI-TRAFFIC-DEEP and AI-CROWDDEEP to provide a complete and configurable solution.

As for the vehicle, the solution allows you to count and classify vehicles that pass through a virtual sensor in a given direction. There are three classes of vehicles detected: motorcycles, cars and trucks. In addition, AI-SMART-CITY estimates the color and average speed of vehicles, generating an alarm where this speed is higher than a configurable threshold, determining traffic density in real time and monitoring vehicle flows via the origin-destination matrix .

AI-SMART-CITY can also be used to detect dangerous behavior or anomalous situations on the road in real time, such as wrong-way vehicles, stationary vehicles, U-turns, pedestrians staying in prohibited areas, or queuing.

AI-SMART-CITY provides also advanced people analysis features in both indoor and outdoor environments. The solution allows you to estimate the number of people present within one or more areas of interest and counts the people crossing a virtual line. Furthermore, it is possible to generate an alarm in case of overcrowding situations, gatherings of people, loitering or excessive permanence of people in an area.

AI-SMART-CITY is the solution for public administrations that decide to exploit artificial intelligence for people analysis and enhance the functionalities of their systems, also existing video surveillance systems. In fact, the solution provides the detection of anomalous behavior, such as people who remain in an area for too long (loitering), or people who are still (for example for the detection of acts of vandalism, such as graffiti). It can also be also used to identify potentially dangerous situations such as gatherings or overcrowd.

AI-SMART-CITY finally allows us to respond to green energy needs, thanks to the possibility of automatically personalizing and regulating the light intensity of the streetlamps placed on the streets. For example, at night we can impose a low brightness where there are no people and a maximum intensity where there are people passing by. A similar application can also be adopted within public offices, in order to provide for the switching on and off of lighting or ventilation systems.

AI-SMART-CITY is designed to meet the needs of cities in monitoring and managing vehicles. The solution provides functionalities to analyze vehicular flows, through vehicle counting and the use of origin-destination matrices, as well as estimating the average speed of vehicles, allowing you to identify the routes in which the average speed is higher. The solution also allows you to detect potentially dangerous situations on the road, such as traffic jams, a vehicle going the wrong way or making a U-turn or even the presence of a pedestrian on the road in prohibited areas, such as emergency lanes.

AI-SMART-CITY can also be used for monitoring tunnels or in motorway environments. Ultimately, the solution allows you to collect the data necessary to make informed decisions both to improve road safety and the livability of cities, and also allows you to optimize vehicular flows.

AI-SMART-CITY is a multi-platform application, and can be installed on more than 20 different hardware platforms including cameras (edge), small processing systems (embedded devices), workstations and servers.

Edge

Embedded

Server

If you have chosen to use a workstation or server-based solution, please connect to our system design support portal: AI-DESIGNER.

Events can be sent to external servers using over 20 different mechanisms, which include third-party VMSs, standard protocols (such as HTTP, FTP, MODBUS and MQTT) and also A.I. Tech proprietary protocols, which allow the notification of events to the dashboards of A.I. Tech. More information via the following link.

Installation constraints

Video tutorial

Online help

Flyer