Video telematics paired with AI will revolutionize construction safety

By Matthew Ison, Head of Sales for North America at VisionTrack

The construction industry is one of the most dangerous industries to work in. It has the second highest number of occupational deaths, and recent data identifies vehicle collisions as the second most common cause of fatal construction injuries. However, video telematics is providing construction professionals with the visibility and data insights to ensure high levels of safety are always maintained, revolutionizing construction health and safety.

While there are safety protocols in place to minimize the risks involved with heavy equipment, it’s hard to eliminate risk altogether. This is especially the case when regularly dealing with moving parts that create a heightened possibility for unforeseen risk, which in turn places additional burden on operators and construction professionals to manage these effectively. 

Video telematics’ ability to provide real-time data for analysis, offers enhanced visibility to support risk detection and prevention, as well as aiding future risk planning.

Covering all bases

When it comes to complex and heavy machinery that require advanced maneuvering, even the most experienced operator can find it difficult to detect a sudden, unforeseen or previously unknown risk and respond quickly and efficiently in the moment. 

Here, a combination of cameras and sensors give drivers a complete 360-degree view around the machine via an in-cab monitor, a view they would not previously have had. Front, side and rear cameras ensure equipment is covered from all angles, while corner, side and rear sensors warn of nearby people – such as workers, pedestrians or cyclists – especially in blind spots. This system can also be linked to specific driving maneuvers, so a driver can view the appropriate camera when turning, for example, the left-hand indicator or reverse gear links to the correct camera angle when engaged.

Added intelligence

Going forward, artificial intelligence (AI) is what will set the next generation of video telematics apart as it becomes increasingly embedded in camera hardware. 

The latest intelligent detection cameras adopt deep learning technology to provide real-time insight. This enables the cameras to identify and track vulnerable people, making it possible to trigger an alert to the driver, especially when manual detection is difficult due to blind spots. The technology can even activate an external, audible alarm when someone enters virtual exclusion zones around the machine, thereby supporting the operator by alerting those around of the risk while preventative action is taken.

No room for complacency

With traditional proximity sensors, alerts can be triggered by inanimate objects, so there is a risk that drivers become complacent and take less notice of an alarm because of false positives. Here, AI-enhanced smart cameras make all the difference. The AI-powered detection camera can locate people up to 20 metres away in real time, establishing the severity of risk depending on their proximity to the equipment. With such high levels of accuracy and range, it provides the driver with increased reaction time and reduces the possibility of collisions.

Non-stop visibility beyond the construction site 

Vehicle and plant camera technology is no longer a simple driving aid. To get the most value out of the data captured, forward-facing cameras and mobile digital video recorders (MDVRs) can now be connected 24/7 to a cloud-based platform, such as VisionTrack’s NARA (Notification, Analysis and Risk Assessment) platform.

Recorded footage and supporting data of any collision, near miss or harsh driving event is uploaded and can be viewed almost instantly. This provides plant or fleet managers with virtual eyes on the ground, enhancing visibility and control of equipment on site or on the road. 

You may also like:

The NARA platform uses deep learning algorithms to process hundreds of hours of driving footage produced by typical construction operations to remove false positives with precision accuracy within a matter of seconds. This means action can be taken immediately, and when it comes to safety every second counts.

Cloud platforms such as NARA can improve safety compliance with additional safety features like the occupant safety rating that monitors the wellbeing of the driver and can even alert emergency services to severe distress or injury following a collision – providing life-saving minutes to operators. 

Telematics and driver behaviour 

Everybody in the telematics industry talks about driver behaviour and being able to determine how a piece of equipment is used or treated effectively. However, with traditional telematics you might see that one piece of equipment has had three harsh breaking events, but without more information this is just general trend information. 

Video footage combined with telematics data provides important context to the recorded incident, unlocking new value from the telematics data. This allows construction managers to gain true visibility into the data to take targeted and effective action. If a video shows someone doing something risky that is linked to the event, then a more meaningful intervention can be made to improve safety. It also allows driving styles to be measured and analyzed, so steps can be taken as part of a wider safety strategy and training initiative.

Construction businesses can also monitor compliance using captured video and data. Almost any activity can trigger a recording by combining video telematics with event tracking on the equipment. 

There can be no doubt the adoption of a compliant video telematics solution has an important role to play in meeting construction industry safety requirements.

Telematics alone can only go so far to prevent safety risks, as it offers only a vague image of the level of safety, with out-of-context incidents and potential false alarms. When real-time video and telematics data are combined in AI-powered video telematics, we have 360-degree visibility not only into activity on the ground but more importantly, into the digital data, to make telematics data more actionable.