Review technical information and learn more about how to deploy, migrate, and build custom solutions for AIRA Server.

AIRA v5 is installable on any commercial off-the-shelf (COTS) PC or server running Windows 2000 or higher. AIRA v5 and AIRA Explorer (AE) are based on a client-server model and Micrsoft's .NET architecture. AIRA v5 communicates with the remote client, AE, via webservices and XML based protocols over HTTP.
The AE client can be used on the same computer as AIRA Server for a stand-alone system. The server can be connected to a LAN and AE will allow users to connect from other computers. If the server is WAN connected, authorized users may even use AE to connect from across the internet.
AE can connect to numerous sites simultaneously since each connection need not send live video until a feed is requested. However, the client will continue to receive health and intelligence information - network errors, equipment status and AIRA generated alarms - so that a live feed can be pulled up when an event of interest occurs. Because the server is not streaming live video to all clients, bandwidth can be kept to a minimum no matter how many users are listening for events. Typical bandwidth requirements for a connection between AIRA Server and AIRA Explorer are 100Kb/s when streaming live video down to 1 Kb/s when AE is listening to alarms.
AIRA Server is completely scalable. AIRA's many-to-many client-server model allows multiple physical systems to be used in an array of AIRA Servers, creating a unified experience for the user. It is recommended, when planning your system, that you plan on approximately 1 server with at least 1 GB of RAM per 16 cameras. It is also recommended that you have at least 2 GB of storage space per 1 day of footage you would like to store per camera. For example, if you were setting up a 32 camera system and wanted one week of full historical footage saved, you would require two servers with about 250GB of storage space each. This storage can be provided by local hard disk, a RAID configuration or even a storage-area-network (SAN).
AIRA can connect to analogue and network devices. Even a combination of analogue, network, fixed and PTZ cameras can all be connected to one server.
For a list of supported video devices, please visit the Device List
section.
Installation of AIRA Server is fairly straightforward. The prerequisites are Windows 2000 or higher, .NET framework and IIS. For installation, connectivity to Aimetis.com is required.
Once AIRA has been installed and registered with aimetis.com, you can begin to use the AIRA system. In order to configure your AIRA server you must set masks, policies, schedules and user access. AIRA has extensive user access leverls in order to offer proper levels of security on a per user basis.
AIRA Server runs the following core services:
AI CompressedMovies - The service which does searching. It finds the matches & creates the video. It also converts .AIRA search results into .MPEG search results and plays.
AI IntelSend - This streams live time graph information from the server to AE. It also notifies AE when a PTZ camera has moved, an alarm has occurred, or the policies have changed.
AI NetSendHist - Responsible for historical video streaming and historical .JPEG creation.
AI PTZ Multiplexor - allows multiple PTZ cameras to be connected to the same COM port. Basically, it is for sharing the resource (serial port)
AI Tracker - The core tracking process. It does the CV analysis, saves video, and does live streaming.
AI Actions - performs the user's specified actions when an alarm occurs (relay, siren, email, FTP, WAV, speak, ...)
AI Cleaner - deletes old footage to make sure we don't run out of disk space
AI IP - Gets the IP address of aimetis.com and writes it to a file so that Tracker can get this information in a non-blocking manner.
AI Rebooter - reboots or shuts down the computer when the user requests this via the admin menu
AI Scheduler - synchronizes access to the CPU to improve performance
Configuration of all AIRA Server's settings can be done through AE in easy to use familiar look-and-feel dialogues. Image masks are drawn directly on recent images from the camera they affect. More advanced fine tuning is available through direct editing of configuration files.
AIRA can be extended and integrated with other technologies such as access control, building automation and other devices.
The AIRA v5 Software Developer Kit (SDK) is available for download. Included with the SDK is a help file with a full description of the AIRA API and 3 sample projects to demonstrate how the interface is used.
The SDK might be used to interface your proprietary front-end UI with AIRA's intelligence and video data, or to interface with the access control system to lock down areas based on activity spotted by AIRA, for example.
AIRA user guide (PDF) can be downloaded here.
AIRA learns how to classify objects over time using a Neurel Network. If the camera is moved, make sure you reset the Learning by right clicking on the camera on the site map.