How AI Will Extend the Security Business
Some of the most interesting trends we're following in the smart security sector involve the use cases for artificial intelligence and machine learning. Because we're still in the early days of these technologies, there's misunderstanding about what AI and ML will mean for these applications. We see four important opportunities taking shape. For manufacturers and security resellers, the time to understand them is right now.
Opportunity at the Edge
According to a survey conducted by ProVigil, some 40% of businesses have changed their security strategy due to COVID-19. The pandemic spurred a need for physical security systems of all sizes to quickly become much more sophisticated. New requirements, ranging from facemask alerting to proximity monitoring and touchless access, have spurred dramatic leaps in system intelligence. To help meet these and other emerging security needs, Edge devices, with their clear advantages in speed and security, have become more important than ever.
A new generation of "AIoT" chipsets designed for AI in edge-facing deployments is making this feasible, and represents the fastest growing IoT platform area. These Edge devices have enough processing power to perform highly complex commands, and enough memory to learn, remember and even act with autonomy. For manufacturers seeking to differentiate their smart security products, the Edge opens up a brand new ballgame.
One of the limitations of legacy systems was the need for outside (read: human) intelligence to evaluate and specify modifications. Once deployed, these systems are of course extensible, but in practice most of them stay relatively fixed. Until now, it took a human integrator to decide how a system could (or should) be extended and when. Today, the system itself can make recommendations, by understanding changing conditions and suggesting pertinent products.
For example, an AI-driven system that can track the local weather can also alert homeowners in advance to an increased risk of flooding or freezing, recommending devices that can prevent costly incidents. Because many security systems include devices from multiple manufacturers, interoperability will be key to these emerging AI concepts. The newly-formed Home Connectivity Alliance will help oversee compatibility and unlock rich new features for end users.
While already well-entrenched in public and industrial buildings, we see AI-driven predictive diagnostics becoming mainstream even in residential deployments. Many use cases are obvious; users need to know when it will soon be time to change a battery or replace a sensor, particularly in security applications, where downtime is not an option. Beyond the obvious benefits of reactive reminders (a battery is almost dead, etc.) predictive analytics can draw from a broad range of data to provide future-facing insights that can reduce costs, enhance safety and improve occupant experiences.
The subject of AI-driven predictive analytics for buildings is so topical that the Continental Automated Buildings Association (CABA) has just released an in-depth research study on the subject. The study has intriguing recommendations for OEMs, software providers, service providers and building/property managers. We'll be on the lookout to see which forward-thinking companies will run with them.
Improving the User Experience
Easy is good and easier is even better. With that in mind, it would not surprise us to see manufacturers using AI and ML to simplify automation routines that many end users would want, if they only knew how to create them. Nest pioneered the way here with thermostats that learned how cool or warm you liked your room and automated the climate based on demonstrated behavior. We see more security systems taking this concept further.
How many other commands could be automated in a smart home just by observing and asking? How about a lighting mode that learns the kids' bedtimes and offers to keep their lights off afterward? Or smart lock that knows you're home for the weekend working in the yard, and doesn't lock itself up again until you're done? These are simple routines that are often too difficult to program for many consumers. They don't have to be, and they won't be.
We think the benefits of AI and ML will become obvious as smart security systems continue to evolve. We look forward to seeing more of the latest developments in these areas at the upcoming ISC West Expo.
Lew Brown, Partner, bluesalve partners is a consummate team builder, leader, deal maker, strategist and implementer in Consumer IoT, Consumer Technology and Consumer Goods.
# # #