必赢体育

On busy construction sites, people can reqularly find themselves working in equipment blind spots

Healthcare, banking, retail, entertainment, manufacturing 鈥� almost every sector of business is utilizing AI, whether through chatbots and virtual assistants, targeted advertising, navigation systems, robotic applications or facial recognition.

To this point, AI systems within the field of construction have been very much on the periphery. While they have made some impact in the world of drone imagery and the generation of 3D models, their use has largely been confined to the back office 鈥� where they do undoubtedly boost the efficiency of tasks such as project management and quality control.

Things look set to change, though, with the development of AI-based systems for object detection that could literally change the game for a .

The business case for reduced risk

Dawn Cirkl, CEO and co-founder of technology company 3rd Eye Robotics, is in no doubt that new technology that can reduce risk will appeal to the industry. 鈥淐onstruction owners want to keep their workers safe and they will do what they can to make that happen,鈥� she says.

In 2023, in the US, 1,060 construction workers lost their lives, and it is estimated that construction is around five times more hazardous to life than the cross-industry average.

In construction, may not be frequent, but when they do happen, they can be catastrophic.

An AI-based detection system can recognise whether a human worker or another object has entered its field of vision

鈥淎ny loss of life is an awful thing for an employer to deal with,鈥� says Cirkl. 鈥淚t can also be costly, in the short-term, with potential payouts, but in the long-term too, with the potential for serious reputational damage.鈥�

Today, safety is particularly high on the agenda for employers, with a large number of older and more experienced operators retiring, to be replaced by younger and less experienced workers.

鈥淲ith less experience in the field, employers want to ensure effective safety measures are in place,鈥� says Cirkl.

Moral and ethical considerations aside, there鈥檚 a strong business case for making the workplace as safe as possible; namely, the EMR (Experience Modification Rating).

The EMR is a measure of how much risk a company is seen to be taking regarding its workplace safety management.

A good (ie low) rating can not only lead to reduced insurance premiums, but also a competitive advantage when bidding on projects, as some clients may prioritize contractors with a lower EMR.

Moreover, workplace accidents, whether or not they involve injury, will always lead to downtime, unplanned expenditure and a commensurate fall in productivity.

Allied to its detection capabilities, 3rd Eye Robotics鈥檚 OmniVai system connects the field and back office through a fleet management portal

Yet, according to Cirkl, contractors do not add the potential cost of accidental damage to equipment to any project bid. 鈥淩arely, if ever, does anybody do this,鈥� she says, 鈥測et it is a direct hit to the bottom line鈥�.

Is the answer in AI vision technology?

鈥淒espite the widespread adoption of backup cameras on heavy equipment, operators often don鈥檛 use them as intended. One key issue is the placement of the display. Many systems require operators to turn their heads or shift focus away from their work, making them impractical in real-world conditions. A display that demands their full attention can be more of a distraction than an aid for operators.

Cirkl says, 鈥淭raditional backup cameras and radar systems rely on the operator to monitor the display constantly, rather than proactively drawing attention to hazards. Over time, this leads to complacency, as operators revert to relying on instinct and direct line of sight, rather than checking a screen that may or may not display a critical warning in time.鈥�

Could AI-based technology then be something businesses turn to, to boost worksite safety? And, perhaps more importantly, would they trust it to keep people and equipment safe?

鈥淚 think people really do trust AI,鈥� says Cirkl. 鈥淓verybody basically understands what it can do now; they see it and use it in their cars and in their homes, on their mobile phones. They absolutely get the concept, but there is still some reluctance to put it to work on construction sites, particularly for collision avoidance.

With 3rd Eye Robotics鈥� OmniVai system, each of four on board cameras can be individually configured for different detection parameters

So, is price the issue?

鈥淎t first glance, yes. But perhaps it鈥檚 also a matter of understanding that there are significant limits to what cheaper object detection systems can do.鈥�

Cirkl believes AI can move far beyond those limits. She says, 鈥淔or AI to meet its potential in terms of collision avoidance, it has to learn about different objects. Basically, it has to be taught the difference between the objects it must avoid at all costs and the objects a machine might actually need to get close to.

鈥淭his technology is simply not the same as a backup camera or a 360-degree camera 鈥� it鈥檚 an extremely powerful and rapidly developing solution with an important role to play in construction safety.

鈥淎nd where this technology is going is incredible. What AI does is allow you to contextualize the environment around you.鈥�

To put that into perspective, 3rd Eye Robotics is constantly teaching its system to recognize and differentiate between specific items that it might expect to detect on a job site, such as people, other machines and trucks.

鈥淚n terms of detection,鈥� says Cirkl, 鈥渢he AI is doing much of what our own eyes do - scanning the environment for objects of interest and notifying the operator when it鈥檚 relevant. The big advantage of having the AI do this is that it can track multiple cameras at once and never lose focus or stop paying attention.鈥�

Diminishing the fatal four

Dawn Cirkl, CEO and co-founder of 3rd Eye Robotics

AI systems are already extremely good at detecting people and machines, helping reduce the risk to site workers of being struck by a machine or caught between a machine and another object.

These are two of the most common hazards for workers on construction sites, half of what are somewhat gruesomely known as the 鈥榝atal four鈥�.

The other two are falls and electrocution 鈥� often from coming into contact with power lines.

The power lines risk is the next big challenge for machine-based AI, according to Cirkl, who says, 鈥淲e鈥檙e already working on some of the algorithms that will allow us to see those power lines from the ground. This is an exciting evolution for AI, and we recognize how important it is to ensure our systems are future-proofed and able to absorb these developments when they arrive.鈥�

With all operator assistance systems, there is the danger of technology taking away 鈥� or at least diminishing 鈥� operator decision making.

Cirkl, who is acutely aware of this, says, 鈥淪ome of our clients want to be alerted to everything, but I point out that this is where complacency and alert fatigue can creep in.

鈥淚t also becomes more likely that operators will ignore alerts if they are endless, repetitive and often unimportant.

鈥淣ow systems need to work more in partnership with operators 鈥� more like cobots [collaborative robots], telling the operator, 鈥楬ey! You need to pay attention here鈥�.

鈥淲hat the tech needs to do is buy you those precious seconds 鈥� to know that you鈥檙e about to hit something, so you can do something about it. That鈥檚 huge.鈥�

OmniVai and beyond

3rd Eye Robotics developed the OmniVai vision system to alert operators to hazards, through both visual and audible prompts.

The OmniVai system can be 鈥榯aught鈥� to differentiate between different types of object that might be encountered on a construction site

The system is machine agnostic and can be adapted to all sizes and types of off-highway mobile machinery.

Each of the system鈥檚 four cameras can be configured to match the working environment.

As people and objects enter the working radius of a machine fitted with the OmniVai system, they will be detected and the operator alerted.

According to 3rd Eye Robotics CEO Dawn Cirkl, 鈥淥mniVai is highly configurable from the cab, in that operators are able to choose, camera by camera, which objects they want the system to detect, as well as adjusting each camera鈥檚 detection and alert distance.

OmniVai uses algorithms to 鈥榣earn鈥� about and differentiate between different objects that could be encountered on site.

Cirkl explains, 鈥渦nlike traditional systems that often require new hardware for upgrades, OmniVai is built differently. Its controller uses powerful GPU (graphics processing unit) computing, allowing AI algorithms to evolve and improve through software updates alone. This means operators can keep the same controller while the system continuously advances鈥攅liminating the frequent hardware replacements required by non-platform solutions.鈥�

A video demonstrating the potential safety benefits of the OmniVai system, can be viewed here.

This article was produced by 必赢体育 Content Studio, in collaboration with experts from

--------

All images courtesy of

--------