One of the main purposes of the new AVS-01 (Alarm Validation Scoring) standard is to provide law enforcement with consistent and more informative calls for service for intrusion alarms. The intrusion alarm levels in the standard are designed to provide a more accurate description of the event, allowing law enforcement to better decide what resources to dispatch.
AVS-01 alarm level classifications are:
- Level 0 – No call for service (event canceled/cleared)
- Level 1 – Intrusion alarm with no additional or limited information
- Level 2 – Intrusion alarm with confirmed or highly probable human presence with unknown intent
- Level 3 – Intrusion alarm with confirmed threat to property
- Level 4 – Intrusion alarm with confirmed threat to life (non-user initiated event)
While AVS-01 was written to be technology agnostic, technology will certainly play a significant role, as will the central station operator who is the final arbitrator of the alarm disposition prior to making a call for service.
Applying human sensing or human detection technologies to intrusion and other alarms systems can greatly increase the capabilities of these alarm systems in detecting whether a human is present within a protected area.
Technology, especially artificial intelligence (AI), will assist automation platforms in determining an initial alarm level. It is true that AVS-01 does not reference any level of response by law enforcement to the various alarm levels and each jurisdiction will decide how to respond to the AVS-01 alarm levels. But, in discussions with public safety and especially with public safety members on the AVS01 committee, the knowledge that a human was on site made a significant difference in how they handled the response.
The potentially elevated response could be the difference between an apprehension or not. The difference between Level 1 and 2 is substantial. With Level 2, the “confirmed” human presence is straightforward — the operator sees, hears or has electronic confirmation that a human is onsite, e.g., “unauthorized opening with a burglary,” or someone the operator called confirms a person.
The “highly probable” aspect of Alarm Level 2 is where it gets interesting and is where AI and other technologies can have a real impact.
Applying human sensing or human detection technologies to intrusion and other alarms systems can greatly increase the capabilities of these alarm systems in detecting whether a human is present within a protected area. These technologies could include various types of radar, infrared sensors, imaging for human patterns, WiFi sensing and many more.
Here, AI and these technologies can report the “highly probable” existence of a human without the operator seeing or hearing the person. This will expand the use for Level 2 and, as confidence grows within the law enforcement community, will lead to better and appropriate responses and more apprehensions.
As we consider applications by AI and other technologies to Levels 3 and 4, we can see more accurate data being made available to the central station operator. Video analytics can report gun shots, and along with additional data, such as rapid crowd movement or perhaps audio of screams or hollering, this could initially present to an operator as an Level 4. Another Level 4 example, weapon(s) present and being brandished in a “threatening” way.
Other instances could be fighting detected through analytics or an intrusion alarm with the homeowner known to be onsite and many more. For “threat to property,” video analytics could detect burglary tools and entry to a protected property. Audio analytics could detect the breaking of glass and voices. These could be presented initially as an Level 3 and with all alarms the central station operator takes appropriate steps to confirm the alarm level.
As mentioned previously, the central station operator is a key participant in the handling of intrusion alarms, ultimately making the final determination of the alarm level prior to making a call for service. However, as AI and other technologies play an ever-increasing role in assisting in the alarm level classifications, these operators will be armed with more accurate data as they engage with emergency command centers (ECC)s.
This in turn will allow law enforcement to better manage their responses and resources. A win for law enforcement and for the security industry.