Data is the new oil. The adage, coined by the UK mathematician Clive Humby in 2006, refers to how, like oil, data must undergo a refinement process before it can be put to real use. It is the raw material capable of powering innovation. For this reason, data is a strategic asset for any organization today. It promises to enhance every aspect of business – from customer experience to product design to sales strategy. 

Massive amounts of this data come from the plethora of CCTV and IP cameras operating throughout enterprises and public arenas. Adoption of these cameras is being driven by a growing interest in smart cities, IoT, cloud, and industry 4.0. At the same time, advancements in AI have improved the ability to extract actionable insights from this raw video data – to process it and refine it – leading to a surge of opportunities to develop smarter applications and more innovative use cases. 

One such advancement is intelligent video analytics, a set of computer-vision-based AI technologies that uses deep-learning neural networks to analyze videos and “learn” to identify objects, people, activities, emotions – in real time or post facto. This paper explores six key use cases of intelligent video analytics across industries such as retail, healthcare, and manufacturing. 

Why regular analytic techniques fall short

Video analytics has huge growth potential, but it is still an emerging field. A 2018 IPVM survey shows more than half of the companies polled use video analytics sparingly or not at all, and fewer than one in five report using them often. 

Most enterprises today use some kind of data analysis to make key decisions, but they may be missing out on valuable insights by not exploring video analytics. AI-powered video analytics can capture more data from day-to-day business operations, leading to more insights and more informed business decisions. 

For example, a retailer might traditionally use PoS data to learn about customer behavior, restricting them to transaction statistics and sale logs. AI-powered video analytics can reveal how customers interacted with the entire store – time spent in each section, average wait time for checkout, even interest in products they liked but didn’t buy.

The manufacturing floor can be dangerous for workers, and supervisors can have a hard time identifying potential anomalies. An AI-powered system can help monitor video feeds and trigger an appropriate action based on insights thus improving product quality and worker’s safety. 

Thankfully, organizations across industries are gradually realizing the power of intelligent video analytics. Last year, the video analytics market size was valued at $4.10 billion and is projected to reach $20.80 billion by 2027, growing at a CAGR of 22.7% from 2020 to 2027.

Deriving value from intelligent video analytics: Six key use cases

Traditionally, increasing security threats and the need for advanced surveillance have driven demand in the video analytics market. Recent advances in AI and machine learning, big data, edge computing and specialized multi-spectral camera hardware have increased adoption of AI-powered video analytics to move beyond providing basic security and surveillance. These advances have increased adoption of AI-powered video analytics to move beyond providing basic security and surveillance to strengthen operations and improve customer satisfaction.

Other practical use cases for video analytics include:

1. Ensuring patient and staff safety in healthcare
   Healthcare providers are using video analytics on site to track patient inflow, manage wait time, and monitor emergency cases in real time. They’re also being used off site to monitor patients on home care programs, which has been especially useful for responding to COVID-19 pandemic. Multi-spectral edge-based cameras have made contactless diagnostics feasible, thereby protecting the healthcare staff and others around.
   
   
2. Enabling retailers to improve customer experience
   Video analytics uncover hidden trends. Live video feeds provide insights on customers’ buying patterns and product interests, and help retailers market better (new layouts, repositioning products on shelves). Video analytics can predict customer interest based on the actions performed by particular groups in the store, helping retailers address customers’ unstated asks.  Retailers can also manage their store-level inventory in real time based on the traffic inside a store which can drive superior store management.
   
   
3. Achieving operational excellence for manufacturers
   Product-tracing powered by video analytics provides insights on manufacturing processes, preventing bottlenecks and delays and thereby increasing efficiency. Every component or assembled product can be tagged with a digital birth certificate, which communicates with video feeds and IoT sensors. Data can be traced back to the tag and used to make adjustments that would otherwise lead to costly recalls. AI algorithms can “learn” to perform tasks, such as provide anomaly alerts and optimizing processes. Intelligent video also supports floor safety by monitoring employee movement and generating alerts for hazardous zones.
   
   
4. Improving vigilance for governments
   City administrations can automate traffic-monitoring and get help with tasks like identifying vehicles, redirecting traffic, identifying public utility spaces, detecting stolen vehicles, managing parking lots, and forecasting traffic volume. Governments can also use video analytics to monitor crowds for compliance with COVID-19 guidelines. The system can alert authorities if there is a large number of people gathering at a place.
   
   
5. Enhancing efficiencies for banks
   Video analytics can detect unauthorized entries into restricted areas, enhancing security for banks. They can also be used to monitor bank operations in real time to provide insights to improve service and increase customer satisfaction. Contactless kiosks/ATMs, which use smart video to detect and respond to gestures, allow banks to provide safer service during the pandemic, another mode to their omnichannel experience.
   
   
6. Reducing manual effort for utility providers
   Advanced video processing and deep neural network algorithms can detect anomalies and leaks in pipes, increasing the efficiency of condition assessments by reducing the reliance on manual assessments.

Unlocking the new business value

Intelligent video analytics have the potential to connect nearly any organization – from smart cities to retail, manufacturing to healthcare – to data and insights that were not possible with traditional methods. Advances in deep learning, IoT, cloud, and edge computing will provide more opportunities to harness the power of that data, and accelerate business transformations. 

In the meantime, ethics and privacy concerns stand in the way of widespread adoption. The rise of explainable and ethical AI will likely help overcome these challenges. Partnering with an expert in intelligent video analytics can help organizations adopt video analytics strategically, ethically, and accelerate their journeys towards intelligent enterprise by infusing AI into their business processes. 

Intelligent video analytics will surely be the next competitive battleground, and forward-looking organizations that become early adopters of the technology will be uniquely positioned to reap its limitless benefits.