Security Operation

Threat Detection & Response

Overview

In physical security, every second matters. By leveraging AI and computer vision, Ambient.ai is able to continuously monitor cameras and detect security threats, which allows operation team to respond faster. The new security operation design help our customers from various verticals protect their sites, reduce risks and save costs.

Role:

Design Lead

Timeline:

2022 Q1, 2023 Q2

Contribution:

Design Roadmap

User Research

UI/UX Design

Team:

1 PM, 3 SDE, 1 UXD

2 Sales Engineers

1 Customer Success Manager

Physical Security Industry

Challenge: High investment💰 Low Impact 📉

The enterprises in the U.S. spend 120 billion dollars per year on physical security protection. However there's no ability to detect or responds to incidents in real-time. Often the security operation starts when the incident already happens…

Computer Vision

Computer Vision is able to understand camera footage, which unlocks the potential to transform security operation from reactive to proactive.

🤔 But how can we bridge the gap here?

what AI sees

vs.

what human sees

Research

The traditional security operation flow:

Who’s involved in this operation flow?

What are they complaining about?

How Might We

automate the operation flow by leveraging AI and computer vision for Daniel and Steve, in order to improve operation efficiency and reduce security costs?

Ideation

To collect deeper insights into user scenarios and brainstorm potential product solutions, we conducted a design sprint, which involves:
• Interviews with CS, Sales, and customers.
• Identifying the new user flow using Ambient product.
• Developing the design for core functionalities.

The unified platform under the hood:

The new security operation flow:

Core Functionalities

🖥

Live Monitoring Cameras

Continuously monitor and detect 150+ threat behaviors by leveraging existing camera infrastructure.

🚨

Raise Alerts

A visual alert with the corresponding surveillance footage will be raised and sent to team members.

📲

Dispatch to Mobile App

Integrated dispatching capabilities allow teams to faster respond with the visual context.

Design Process

Information Architecture

The visibility of live camera views and real-time alerts are both essential to operators. So the monitor page includes the alert feeds and operators video wall. For additional context and details, operator can open up the expanded alert page.

LiveMonitoring

Expanded Alert Page

Components Iteration

Video management system and alerting experience both require heavy interaction between operator and product to decrease "Time to Act".

1.1 Operator’s Personal Video Wall:

A video wall contains all the cameras of interest. The views can also be switched between [Live Mode] and [Recorded Mode].

Video Wall Layout (lo-fi)

Video Wall Layout (hi-fi)

Video Wall with video controls

1.2 Alert Card:

Alert card with details gives operators context for further actions: [view], [dispatch], [resolve] etc.

Old Design

New Layout

New Design

2.1 Video Controls:

Video controls allow operators to easily navigate on timeline and view the corresponding camera footage. We revamped the old video control component to give more flexibilities to operators.

2.2 Dispatch Modals:

Dispatch modal shows the responder list and allows operators to send dispatch from alert.
The old design didn’t provide much details like status and active dispatch. The new design makes it more visually appealing with clear information architecture.