Fire Video Analytic IP Camera (Fike SigniFire)By: Brian Rhodes, Published on Mar 11, 2012
Fire and smoke detection can be a challenge in large, open buildings. Smoke can stratify, dissipate quickly and can drift a long way from the source in large buildings before traditional sensors are tripped into an alarm state. Video fire and smoke detection is positioned as a superior alternative.
One offering, the Fike SigniFire IP camera [link no longer available], has a different design approach to detection than traditional photoelectric or heat sensors. The SigniFire uses a visual image to determine the presence of smoke or fire. In this note, we look at this camera and competitively compare it to other fire detection solutions.
The SigniFire [link no longer available] is a box format, indoor-rated camera equipped with the following:
- 640 x 480 (NTSC) resolution with MJPEG compression only
- Smoke, fire, and motion detection analytics built on-board.
- Two factory lensing options: 2.8mm (82 degree Field of View) or 8.0mm (34 degree Field of View) fixed focal lenses
- 10 lux minimum illumination
- Terminal block outputs for Fire Alarm system loop
- BNC connector out
- [UPDATED 2013] VMS integration: Fike says their SigniFire camera feeds can be viewed by and alarms passed to third party VMSes via ONVIF by their FSM-IP NVR. Their NVR can also import ONVIF feeds from third party VMS's and run their fire/smoke/motion analytics on them. However, the SigniFire cameras are not ONVIF conformant, but the NVR is, which is strange, adds cost and complexity.
Here is an image of the camera:
Claimed Minimum Detection Performance:
- Fire: 1 foot fire plume at 100 ft
- Smoke: plume verification at 100 ft
Pricing for the SigniFire is $3,000 USD list price per camera including lens.
The video below is an example of Fike's visual detection system in use:
The biggest advantage of this product is the multipurpose ability to detect both smoke or fire at distance in a single unit. Traditional detectors need to be hung close (within 30') to an ignition or smoke source in order to alarm, while video detection is able to alert to a fire even when the camera is hung a considerable (100 ft) distance away. The detection analytics are contained on board the camera, and the unit can be integrated into most fire alarm systems via the included dry contact outputs. While the camera has SD resolution that may result in poor surveillance quality, this resolution provides a very precise method of determining when and where smoke or fire has occurred compared to nonvisual sensors. For large, open facilities like warehouses, stadiums, hangars, or industrial plants, the video detection technology permits earlier alerting of an alarming condition than traditional sensor technology.
A number of drawbacks exist:
- The list price of $3,000 USD per single camera is significantly more than traditional photoelectric smoke detectors selling for around $150 USD.
- It may be used with 3rd party VMS but only via Fike's own NVR.
- Multiple cheaper detectors can be installed for the single price of a SigniFire camera.
- This technology is not intended to replace a traditional fire alarm system. We consider this a prudent restriction, due to the redundancy and fail detection that a traditional multi-sensored zoned detection system possess over a single point of failure in a camera.
- The SigniFire camera is only UL listed as a fire detection device when used with Fike's alarm panels. The output relay in the camera is connected to a fire alarm addressable module in the following manner:
The pairing of fire and smoke detection with cameras is not uncommon, but other options have marked differences. Products like Vumii's EyeSec FD are considerably more expensive and intended for long-range outdoor applications only. Several products, including those from DTEC, are appliance (DVR) based analytics intended to be used with new or existing analog cameras. Because the analytics are sold in groups of channels, the overall cost per detection channel might be lower but the total acquisition cost is higher.
Other analytic products, like InnoSys [link no longer available] and ASL Vision, commonly only address a portion of the total equipment solution and rely on quality of the associated camera for effectiveness. When implementing a system like the above, the performance of the system is greatly affected by the quality of the camera it is associated with. The integration of an alarm event into a fire panel remains a problem solved by purchasing the analytics installed onboard additional hardware modules.