LAS VEGAS – Technology manufacturers from around the world gathered in Las Vegas for the annual CES show to showcase the next generation of devices and systems that will revolutionize the way we live, work and play. From autonomous vehicles and advanced mobile communications, to the latest entertainment gadgets, the show featured a wide swath of technology innovation for myriad industries, including security.
While the foray of CES into security began with home automation and its eventual integration with residential security systems that we know today as the Smart Home, the show now offers a glimpse into other significant tech trends – Artificial Intelligence (AI), Machine Learning, the Internet of Things (IoT) and others – that are now impacting various other sectors the security market, such as video surveillance and access control.
Here’s an inside look at some of the companies I visited and the cutting-edge technologies that caught my eye on the show floor:
Smart Home
Alarm.com earns Product of the Year: I got a chance to catch up with Matt Zartman, Director of Communications for Alarm.com, who showed me the company’s new “Smart Signal” System, which lets a user easily and securely cancel a false alarm with a single command, with no need to talk to a central station or remember a passcode. The user simply holds the cancel button in the app to let the monitoring station know that no further action is needed.
Smart Signal was honored as “Security Product of the Year” at the Consumer Technology Association's Mark of Excellence Awards. Read more about Smart Signal at www.securityinfowatch.com/21040000.
In addition to Smart Signal, Alarm.com introduced Wellcam as “The Smarter Video Solution for Independent Living,” allowing family members to respond to alerts from an elderly family member. Both are very forward-thinking solutions.
ADT Pulse redesigned: ADT’s new Command Panel & Control Platform (ADT Command for short) is a complete redesign of ADT Pulse, first released in 2010. Command supports more than 250 devices, ranging from thermostats, lights, doors, carbon monoxide detectors and smart light bulbs.
At the core is a seven-inch wireless touchscreen command panel with integration to Amazon Alexa and Google Voice Assistant to enable dynamic scenes and rules, so users can use voice commands to run several routines at once.
Add dual path broadband and LTE cellular communications with two-way encryption, a redesigned video doorbell, the new ADT Go App (Apple or Android) and interactive crime maps, you have a smart home solution to support one of the largest monitoring services.
Read more about the system at www.securityinfowatch.com/21040493.
Resideo is ready: Now that it is a fully independent spin-off from Honeywell, Resideo has established itself in the growing smart home with such staple products like indoor/outdoor cameras, boiler and HVAC controls, temperature/humidity comfort control, water leak detection and cloud applications to monitor this suite of sensors. The company made several product introductions at CES.
The company unveiled the Honeywell Home T-Series Smart Thermostats, which will be available this spring. At launch, the thermostats will integrate with Amazon Alexa and Google Assistant. The T9 Smart Thermostat will be available direct to consumers; while the T10 Pro Smart Thermostat will be available via professional heating and cooling contractors. The wireless smart sensors that come with the T9 and T10 include people detection and geo-fencing features.
Resideo also introduced the SiXCOMBO detector, a two-way professionally installed and monitored wireless combination smoke/heat and carbon monoxide (CO) detector, as well as an integrated, mobile home automation control panel in a tablet form – providing controls, security monitoring, video surveillance, perimeter status and comfort controls.
Ring still going beyond the doorbell: The new Door View Cam from Ring replaces your peephole, and is capable of sensing when people are knocking on your door. It is easily installed in place of the existing door viewer or peephole and does not require drilling or permanent modifications to your home or door.
For Ring Alarm full security systems, the company introduced a full line of sensors to connect to the Ring Alarm Base Station, including a smoke and CO listener, flood and freeze sensor, dome siren, First Alert Z-Wave Smoke/CO Alarm and smart locks from providers including Kwikset, Schlage and Yale.
Ring's Smart Alert algorithm is an internal program that runs inside of the device. It analyzes the traffic patterns of detected motions, looking for repetitive patterns. When it detects such a pattern, it essentially "sleeps" your device for a certain period of time, cutting down the number of motion alerts from the same set of stimuli. The Smart Alert controls fine tune the Smart Alert Algorithm based on your monitored area, just like the menu of an IP camera with video analytics.
Schlage among the Alexa-enabled: Many home security locking systems with Alexa were spotted at CES, such as the Schlage Encode Smart Wi-Fi Enabled Deadbolt, the first Wi-Fi connected smart lock that does not need a hub. It can store up to 100 unlock codes that can be permanent, temporary or active during certain times. With the lock’s Alexa skill enabled, users can lock and unlock the door remotely and check the lock’s status at any time.
Artificial Intelligence
It was clear that the AI leaders at CES 2019 recognized the clear need for a continuous stream of data to train on, leveraging mature AI processes such as object detection, classification and natural language recognition. Here are some of the specific applications that may end up being relevant to residential and commercial security:
Ambarella – for me, “Best in Show”: Although Ambarella produces a significant range of processors often already in the smart devices or visual imaging systems you use already, its CV25 demonstrations were, for me, best in show.
CV25’s Computer vision features license plate recognition, retail behavior analysis, face recognition, and object classification includes pet and person recognition, package detection, voice-operated command functionality and intelligent alerts.
One of many visual demonstrations showed: A busy traffic intersection, a freeway, a crosswalk with pedestrians, a vehicle moving down a narrow street with parked cars, pedestrians and road hazards and a freeway no-stop toll collection arch. Every vehicle was rendered smoothly with object classification target boxes, as well as displays of percentage recognition confidence level, direction of travel, vehicle license plate and make of car or vehicle – even at speed on a six-lane freeway with multiple vehicles abreast; and even stopped at an intersection, where the density and quantity of the vehicles cannot be determined with the naked eye.
One demo contained footage where the streets, vehicles, pedestrians, obstacles, buildings and foliage were colorized, in real time, with the same stellar performance. This feature will be essential for both public safety and public transportation. It will even aid forensic review.
As an industry, we are used to relying on edge enhancement and wide dynamic range for visual acuity. The new breed of AI chipsets have gifted us a new method. It is up to the systems integrators and end-users to specify processors like Ambarella and that the camera OEM makes full use of the CV25’s feature set.
Watch my video of the Ambarella demos...at around the 4:30 mark, you will see another visual CV25 demo that looked like a thermal infrared camera rendering of the three of us in front of the display – it was actually a 3D image extracted from a stereo camera that was color-coded by distance. You could actually see the contour of an object I had in my jacket vest pocket; thus, it is not hard to imagine this as a capable, cost effective alternative to security screening for events, casinos and schools.
The visual presentation required some software to render the graphic elements, including object classifications – all made available through open source. To set realistic expectations, the systems integrator and end-user need to balance their software and integration needs.
All of the complex AI processes displayed here were supported inside the processor itself –without the cloud and without reliance on GPUs – all at very low power. However, it is clear that Ambarella has included a suite of features that may not be implemented by your favorite IP camera manufacturer. This is why CES is so essential to executives in our industry – as an educated purchaser, you know what is capable and there will most always be a provider who is using this processor.
Nvidia’s Deep Learning Super Sample process: On the opening night of the show, Nvidia CEO Jensen Huang explained how most computer graphics today use a process of rasterization – whereby objects are rendered a bit unnaturally as collections of triangles, which fails to accurately produce complex nuances of light and shadow.
Enter Nvidia’s new "deep learning super sample” (DLSS) process, where difficult scenes having nuances of shadows, reflections on glass and in water and many shadows are now accurately generated. This opens object recognition to once limited areas of object, traffic and facial recognition for the security industry; in fact, investment bank UBS reports the demand for mass facial recognition and traffic monitoring could be worth up to $5 billion in additional sales for Nvidia by 2020.
Z-WAVE>ME turns users into the programmers: In the Z-Wave Alliance pavilion of partner mini-booths, a relatively small AI application developer showed me something amazing for popular microcomputer platforms like Uno and Raspberry Pi – already embraced by the residential, DIY and small systems security markets.
The solution, by Z-Wave Alliance partner Z-WAVE>ME, simplifies the creation of Z-Wave devices and enables someone to create a small AI system. The DIY “creator” would design, code, test, recode, test again and finally implement jump from design to implementation – using a platform that automatically generates and tests the AI code. That’s an instant, customized DIY system. After demonstrating this, now I am confident I can quickly develop an AI-based water, CO, pool control, AC unit status monitoring solution with built-in notification logic.
Hisense – perhaps a glimpse into the monitoring center of the future: Hisense demonstrated how AI-powered TV algorithms gather information about what you are watching and then automatically optimize the settings to enhance the viewing experience. For example, when the TV’s AI identifies your program as a sporting event, it will automatically adjust the motion rate to give you a HD realistic picture. Time will tell if this technology can be adapted for or adopted within security monitoring center applications.
FLIR’s machine learning…but for autonomous vehicles: FLIR showcased VR and one of the best Smart City visual platforms I have seen – combining traffic status and incident reporting. But the real star of FLIR’s CES 2019 show was an amazing, tiny Automotive Thermal Imaging Sensing System, capable of placement virtually anywhere on a vehicle. The system runs concurrent with its second-generation Automotive Development Kit (ADK) with machine-learning algorithms for object classification. The thermal-vision ADK augments the entire sensor suite and offers the redundancy needed to improve safety in autonomous vehicles.
Voice Assistants
Amazon vs. Google: Voice assistants are being integrated into every part of our daily lives, from home security to industrial “skills” for smart cities like Las Vegas. Amazon recently announced that more than 100 million devices with Alexa have been sold.
Alexa Voice Service (AVS) is integrated with Amazon's e-commerce environment, streamlining purchases. The system can operate as a home automation hub, allowing the user to control heating and lighting systems, streaming media services online, and supports If This Then That (IFTTT), an online service that automates Web-based tasks so that when user-specified events occur, follow-up tasks are triggered and handled: Home Automation.
Key by Amazon is a growing set of services that enable users to experience life without physical keys. Starting in the first quarter of 2019, Ring customers will be able to control Key by Amazon-compatible smart locks using buttons in the Ring app, as well as view live video from Ring video doorbells and cameras.
Key for Business, available now, is an Amazon package-delivery service for commercial and residential properties with compatible building access systems. With Key for Business access via smart fobs, delivery drivers will be able to put packages in secure areas without requiring assistance for entry from building staff personnel.
As much as Alexa was buzzing, “Hey Google” was everywhere, messaging the global leaders approach to having a natural conversation with your devices, to fit your lifestyle, rather than a global library with Alexa devices as the entry interface. Google Assistant will be available on 1 billion devices globally before the end of January 2019. That is ten times more than the 100 million devices Amazon said had access to the Alexa voice assistant.
Drones & Robotics
FirstIZ for first responders: While the security industry struggles with rogue drones shutting down busy and critical aviation hubs, FirstIZ quietly introduced two Unmanned Aerial Systems (UAS) that clearly will set the design standard for future aerial systems. Virtually every safety and security sensor can be placed on either UAS including thermal imaging, wide area stabilized imagery, ground fire detection, combustible and harmful gas detection – with detection even in motion or at speed.
Far better than the anti-drone solutions that are in their infancy, both drones can detect multiple groups of rogue drones and report back to command with an accurate First Person View (FPV). Clearly, FirstIZ products are ideal for aviation, first responders, event security, critical infrastructure monitoring and the gas and oil industry.
Training Robots as first responders: MediaTek had an amazing AI training and “mimic” exhibit at CES, where a woman’s actions in front of a green screen were synchronized by an AI-powered robot, which was able to keep up with most of the poses. This will represent a life-saving leap forward for first responders, when a robot can imitate a first responder’s movements to enter a highly dangerous environment where a human would be at too great a risk.
Security and Safety Things (SAST) Ties it all Together
Bosch spin-off SAST – unveiled at GSX in 2018 – hopes to answer the question: How can the electronic security industry integrate video technology from Consumer Electronics, Autonomous Vehicle, Smart Home, DIY and AI industries? Thus, it only made sense that SAST would have a presence on the CES show floor.
One of the most promising innovations in the Security Ecosystem is the creation of four key services:
1. The first “standardized” OS platform to support virtually any video surveillance camera with transparent upgrades, similar to the way Google Play and App Store handle updates;
2. An integrator portal designed for delivering video services;
3. A world-class app store to purchase video management, analytics and related applications; and
4. A Software developer environment designed to encourage collaboration on the new platform and move the SAST community forward.
Watch my video interview from the SAST booth at CES.
Steve Surfaro is Chairman of the Public Safety Working Group for the Security Industry Association (SIA) and has more than 30 years of security industry experience. He is a subject matter expert in smart cities and buildings, cybersecurity, forensic video, data science, command center design and first responder technologies. Follow him on Twitter, @stevesurf.
