Hanwha Vision discusses the MQTT protocol for AI cameras and the Internet of Things (IoT)

When the Message Queue Telemetry Transport (MQTT) protocol was invented over two decades ago, its creators perhaps didn’t realise the widespread applications it would become central to, across all industries.

That’s because the MQTT protocol has become the de facto standard for sharing messages across connected devices — also known as the Internet of Things (IoT). It provides the means for IoT sensors, including cameras, to communicate with each other across smart cities, smart buildings, and in different verticals including retail, healthcare, and manufacturing.

Why MQTT?

The MQTT protocol is ideally suited for this application because it is an extremely reliable and lightweight messaging transport protocol with minimal network bandwidth and a small code footprint. Using a so-called ‘publish-subscribe’ approach, it queues, shares and relays messages in a highly efficient way, making it ideal for connections between devices hosted at remote locations that have resource constraints or limited network bandwidth.

Many Hanwha Vision networked video cameras work seamlessly with any number of other sensors

It is also based on open standards, so has the flexibility to work with a vast number of devices, in endless applications including street lighting, access control, traffic monitoring, parking management, and environmental quality.  

Thanks to MQTT, many Hanwha Vision networked video cameras work seamlessly with any number of other sensors you have in an installation. Meaning that your video surveillance can now do a lot more than security monitoring — it can have an impact on all sorts of business and city operations. It’s true to say that MQTT and video in tandem bring many smart city, building and retail dreams to life. 

Actionable insights in real-time

The number of IoT devices is rising exponentially. Current predictions state that connected devices will exceed 13.1 billion globally by the end of 2023. That’s creating a lot of data, and a host of opportunities to make organisations more intelligent, efficient, and personalised.

Take, for example, someone walking into an otherwise empty building. Using the MQTT protocol as the means of communication between devices installed in the building, the occupant’s arrival could trigger a range of actions.

When someone is detected to have entered a building, smart lighting could be requested to turn on, and a heating or air conditioning system to make the environment more comfortable for the occupant. Because MQTT is open, it will work with endless smart devices — including a smart coffee machine. So as the individual walks through their smart building to the kitchen, they could even find a fresh brew waiting for them!

MQTT protocol

Video analytics monitor occupancy levels and trigger alerts if more visitors arrive

For busier areas, video analytics could monitor occupancy levels and trigger alerts if more visitors arrive and exceed safe numbers. People could be directed away from an overcrowded space to another, quieter area, automatically, the result of devices communicating using the MQTT protocol and taking a predefined course of action. 

Alternatively, more staff could be needed, with notifications sent to smartphones or tablets to ask them to move to the busier area. This could prove particularly useful in shopping centres or transport locations where the visitor experience is critical to overall customer satisfaction. 

Bringing the ‘smart’ to a smart city

On a citywide scale, more sensors could be used to monitor air quality around a city’s streets. Through MQTT, this data can be connected with live traffic data from cameras and road sensors, to understand whether a build-up of congestion is contributing to a fall in air quality. Traffic could then be re-directed to less busy roads to enable pollution to fall to acceptable levels.  

A fall in air quality might not be caused by a build-up in traffic but instead by an emergency. In this case, having a rapid response, facilitated by devices communicating with one another, can be lifesaving. A sudden fall in air quality could trigger a control room screen to show images from thermal cameras, providing confirmation of a fire that’s producing toxic smoke.

Digital signage and a public address system could then direct the public away from the area for their safety. Not only does this prevent the situation from worsening, but it also allows space and time for first responders to arrive. 

Long-term benefits

Business pioneers are increasingly realising the value of data when planning for the future

It isn’t just immediate impacts where MQTT communication between devices proves useful, but also in longer-term strategic decisions. Business pioneers are increasingly realising the value of data when planning for the future.

Video and IoT devices are rich sources of visual, environmental, audio, temperature, and other data. Bringing all data sources together into a coherent, easy-to-understand interface allows pioneers to make full use of all available insights. 

Understanding store occupancy, footfall traffic

In retail, this will likely take the form of an even greater understanding of store occupancy, footfall traffic in specific areas, optimum staffing levels, and energy usage. Being able to see when stores regularly experience a surge in customers can inform staff rotas.

Store layout can be influenced by footfall and occupancy data (and high-value, best-selling items put in spaces where more people walk past). Even heating and ventilation can be programmed based on how many people are visiting a store — of course, this also improves energy efficiency as HVAC is only operating when, and where, needed. 

IoT plus video data

Data collected by each one remains siloed and hard to use with any regularity and consistency

In a smart city, understanding the flow of vehicles and people throughout a space will help city planners create roads, pavements, and public spaces that work for every citizen, whatever their mode of transport. Busier areas can get more road maintenance, while quieter areas may benefit from additional street patrols.

All smart city pioneers want to make their citizens feel satisfied and secure, and using IoT plus video data is an easy way to see in near real-time what’s happening, to improve response times and tailor a city to its citizens’ needs.

MQTT plus AI: an ideal pairing!

It would be remiss to leave out the role of AI in these applications. Video cameras feature deep learning on top of more ubiquitous machine learning and basic video analytics. More analytics can now be done ‘at the edge’, in the device itself, so having additional data coming in from across the IoT via MQTT communication is invaluable. It can create alerts, trigger actions, and deliver insights that are more contextual, deep, and useful.  

In many ways, MQTT is critical for AI to continue its advance in the video sector because it allows cameras to communicate with other devices. Otherwise, the data collected by each one remains siloed and hard to use with any regularity and consistency.

Bringing MQTT to Hanwha Vision

MQTT has come to Hanwha Visions' full range of AI cameras and with it, all of the benefits and possibilities offered by the communications protocol.