The Costs of a False Alarm
Reducing false alarms pays off.
The mobile surveillance market is quickly growing and rightfully so. Being able to simply place a self-contained camera mast at a site to monitor intruders is an obvious way to improve security at reasonable costs. Furthermore, we see the mobile surveillance applications quickly expanding. Over the last year we have worked to improve the performance of movable masts intended to monitor construction site (to a point that we are helping recognize theft of particular objects), we have worked together with partners to build a mobile trailer intended for rapid surveillance of event sites, and we have collaborated to use edge AI to allow movable camera masts to count and classify traffic. Portable, stand alone, surveillance masts combined with advanced edge AI allow for a variety of applications.
However, despite this big uptake and the amazing use-cases, we still see one common problem with the currently existing solutions: many surveillance companies, particularly in demanding mobile situations, experience a huge number of false alarms. In this post we examine why false alarms are a problem, why the two most common methods to solve this problem are infeasible, and how the problem can quickly and cost-efficiently be solved using a smart gateway.
So, why are false alarms even a problem?
Let’s start by describing the actual problem: Many (mobile) surveillance solutions provide alarms whenever something suspicious happens. These alarms often trigger a cascade of actions:
- When an alarm is raised, a network connection is setup to allow for remote monitoring of the site.
- The video data is sent, over a mobile network (often 4g/5g) to a VMS.
- A human operator receives the alarm and uses the VMS to inspect the images coming from the camera.
- If the alarm is valid (i.e., there is an intruder), the human operator takes appropriate action.
Note that every alarm leads to an increase in energy usage of the mobile surveillance solution, a large increase in data bandwidth, and additional human labour.
Regretfully, most camera systems provide alarms based on simple movement detection (or very poor, early 2000s, machine learning object detection). Thus, on a windy and rainy day, we have seen cases where a single mast produced over 500 alarms in a single night! With so many false alarms the networking costs are immense, the energy consumption drains the masts batteries, and the human operator will surely loose interest. All in all, any false alarm has direct costs associated with it, so reducing the false alarm count is essential.
This should be solved shortly, no?
We often hear that the false alarm issues will be solved shortly: “once we move over to our new VMS or to our new ‘smart’ cameras” the issue will be solved. While these solutions are clearly infeasible for brown-field masts and thus require large investments, even as a future proof solution these directions do not work:
- Relying on cloud based AI, provided by a smart VMS, to effectively support the human operator is still extremely energy consuming and still confronts you with high networking cost. The videos streams still need to be send to the cloud.
- Relying on new “smart” cameras to run AI on the mast itself is a step forward, but these cameras are a) overly expensive while being directly exposed and thus vandalism prone, and b) hardly future proof: most smart cameras come with a single set of AI models that are presented “as is”. If these models fail to provide the performance or functionality you seek now or in the future you are simply out-of luck. In a few years, you will be stuck with poor, early 2020s, AI.
So, while the problem is clearly recognized in the field, the current solution directions are infeasible.
An intelligent gateway solves false alarms and makes your mast future proof.
Together with our resellers we provide a different approach. We run advanced AI models on a gateway in the mast to reduce false alarms. These advanced AI models — which can efficiently be used on supported gateways using our patented AI software — have shown to reduce false alarms by over 99% thus saving the networking and human costs involved while being extremely energy efficient. Running AI on the gateway has many benefits:
- The smart gateway solves the false alarm problem “out-of-the-box”. The gateway can simply be installed and remotely configured to work with your camera’s. This is even true for existing, brown-field, cameras.
- You can use relatively cheap, “dumb”, cameras in the mast. The processing happens on the LTE gateway itself which can be securely housed. And, you will need a gateway anyway.
- You can purchase the gateway through our partners with all the software and necessary hardware pre-configured.
- The gateway’s AI is future proof: it is easily extensible to include PPE monitoring, vehicle counting, and even license plate recognition.
Of course, the gateway provides all the networking functionality you would expect!
Features of the smart gateway
The smart gateway provides a number of sought-after features to optimize your mobile security solution and make it future proof:
- The smart gateway (depending on the model) supports up to 4 camera’s at once.
- The smart gateway provides remotely configurable masking for each camera.
- The smart gateway can be set to run the advanced AI models upon changes in the image: this combination makes the gateway highly effective and highly energy efficient.
- The smart gateway can send alarms to your VMS, but alternatively can send alarms directly to you: you can simply open up a VPN connection to inspect what is happening on each respective camera.
- The smart gateway can easily be upgraded with new AI models: Recognizing driving plates? Recognizing license plates? Counting vehicles? Recognizing empty parking spaces? All of this is easily and remotely configurable.
- The smart gateway works with virtually any modern digital camera. Upgrade your current solution to include advanced AI capabilities, or design a new solution based on simple cameras and highly advanced AI running locally on the gateway.
Get started right away
We are quite proud of our developments on the smart gateway. Starting from a purely technical solution for running advanced AI models highly efficiently, we have advanced to a plug-and-play surveillance solution that we, together with our partners, can offer at scale. You can start reducing your false alarms and making your mobile surveillance solution future proof by contacting our sales directly using our chat, or by booking an online demo.
Do you want to stay ahead of the curve in edge AI? Register for our newsletter.
Nothing beats good documentation
As we have more and more users directly using our platform and the AI manager to create and manages edge AI solutions, it has become increasingly important to provide up-to-date, easy-to-read, and complete documentation. There is so much you can do using the Scailable platform, and immense engineering effort that is saved by using the […]
Apples and Oranges II: The forgotten metric.
In an earlier post we described the performance of the Scailable AI manager in terms of computational speed. While this is an interesting metric, in many ways a solution only has to be “sufficiently fast”; once this is the case other metrics become much more important. Power consumption might be one (and is one we […]
Apples and Oranges: Profiling Edge AI solutions
In this post we examine in some detail a question that we get repeatedly: “How does deployment of Edge AI models using the Scailable AI manager, on Scailable selected hardware, compare to other deployment methods in terms of performance?”. Although the question seems easy, the answer needs a bit more explanation which we try to provide […]