For most of us, Hollywood has painted an elaborate picture around the world of robotics and artificial intelligence. Images of dystopian societies, robo-villains, and Arnold Schwarzenegger's classic line "Hasta la vista, baby!" probably come to mind.

The New York Times recently published a rather inspiring piece A Smarter Way to Think About Intelligent Machines where author, Dr. Ed Finn, challenges readers to "tell better stories about all the everyday ways A.I. is already changing the world."

A better conversation around robotics

This struck a chord for me. Rather than fearing a fictional idea of robots taking over our society, let's create a productive dialogue around the revolutionary technology that allows humans to work smarter, safer, and happier.

Now, let's dive into some of the features that make robots so disruptive to the security space.

1. Autonomous Patrol: the Crowd Pleaser

A security robot without autonomous capabilities is about as useful a Monday morning quarterback. The ability to patrol and re-charge without human intervention is critical to making these robots a worthy investment.

Autonomous patrol starts with localization on a digitally-generated map. Most robots use lidar technology to create these maps and get spatial awareness in their environment. At a high level, these security robots operate similarly to autonomous vehicles: using lidar mapping for spatial awareness, and machine learning and aggregated data to get smarter over time.

This is a big win for the security industry. Automating patrols allows for security officers to spend more time on the managerial and strategic aspects of their role, while still getting accurate data in real-time.

2. Event Detection: the Show Stopper

Anomaly detection and automatic responses typically determine the appropriate applications for a security robot. Here are some common anomaly-types that are important for facilities to track:

• Temperature changes (especially in data centers)
• WiFi access (tracking if a blacklisted device has entered an area)
• Sound detection (ie: breaking glass, shouting, or gunshots)
• People, animal, and object detection (most common form of detection)
• Facial recognition (this is still ambiguous in its legality)
• License plate detection (commonly uses OCR, which isn't always accurate in certain environments)

Pairing an anomaly-type with an automatic response is the ideal recipe for emergency response and facility management. For example, consider an active shooter situation on a corporate campus. If a robot detects the sound of a gunshot, it can automatically alert authorities and campus staff with a time-stamped location, followed by initiating the given organization's procedures for emergency response - such as a total lockdown or sounding an alarm.

3. Real-Time Alerts: the Mic Drop

As you probably know from any group project in grade school or your team's Monday morning meeting, communication is everything. Event detection is only as good as its notification system.

Creating custom alerts keeps teams informed without information overload. In most cases, it's helpful to schedule alerts. For instance, creating an alert when someone is in an area overnight or during restricted hours.

Customizations aren't the only key feature here, but also the method of delivery. For most security teams, the ability to access information on-the-go is critical, so an email or web notification might not always fit the bill. That's why mobile push notifications win the MVP title for delivery method.

4. Live Video Streaming: the Extra Mile

Live video is a handy tool for monitoring a developing event or resolving an issue remotely. It goes hand-in-hand with instant notifications, as mobile access to live video feeds completes the circle for responding officers by providing up-to-date information and perspective.

Recording and organizing live video also helps with event management. Most security robots offer software that automatically categorizes, organizes, and exports relevant event footage to assist with evidence and data collection.

Security robots rely on a stable network connection to transmit and compute data in real-time. Certain applications reveal challenges that engineers must tackle before releasing a robot into the market. For instance, establishing a strong network connections in areas of limited bandwidth (remote facilities) or high traffic (public buildings, densely populated urban areas) can prove to be a test for engineering teams.

5. Concierge Service: the Icing on the Cake

Automating some building management services can be a big cost-saver for businesses willing to invest some initial time and money. Hilton and IBM collaborated to test a robot concierge service that provides guests with information on nearby attractions, hotel info, and restaurant recommendations. Although security robots aren't intended solely for this use, it can sweeten the deal when pitching its value.

Cobalt Robotics provides a multi-faceted solution with a friendly UI that can assist visitors with directions, scanning visitor passes or access cards, and direct communication with a visitor help center.

It's safe to assume that further development of these features will continue in the next few years as the market grows accustom to automated assistance.

6. Outdoor Patrol: the Holy Grail

And finally... what you've all be waiting for: outdoor security robots. Believe it or not, this is the most common request we get during our market research from end users, security companies, and system integrators. Everyone wants an outdoor robot.

The technology isn't quite there yet. Although there are a number of solutions that are already on the market, there are limitations in terms of applications and environment.

However, areas that are more remote or rural demand an adaptable solution. As previously discussed in #1, the autonomous patrol features relies heavily on lidar sensor data. In environments that are unpaved or uneven surfaces (the perimeter of a utility facility for instance), robots can be easily delocalized if a wheel slips or enters a free spin.

Additionally, network connectivity poses a challenge to these robots. In remote areas, it can be difficult to secure a strong network that can handle cloud computing, fast upload/download speeds, and reliable live streaming. In areas that demand a rugged, outdoor robot, this is just as much a challenge as the localization issue. Creating solutions to these hurdles comes at a price, as most outdoor robots currently start at $50,000 annually.

In this category, we have to wait for technology to catch up to our imaginations. Once economies of scale kicks in (we're predicting that will happen in the next 5-10 years), it won't be out of the ordinary to see a security robot patrolling the parking lot of your favorite grocery store or big box retailer.

The Wrap-up: This is only the beginning.

In an industry that's only a few years old, we're innovating at a rapid pace. Current market players are iterating their robots and services to find the right market fit. Expect to see more innovation as technology evolves to become even more mobile.

These chart-topping features are only the tip of the iceberg. In the next decade we will continue to see jaw-dropping security robots start to evolve.