Vision AI Software and Hardware are Working Together to Make the Workplace Safer
In 2019, the US National Safety Council reported that the country spent $171 billion dollars on workplace-related medical bills, affecting over 158,000 workers. This is before the Covid-19 pandemic shook the very foundation of our society, forcing isolation and new distancing mandates globally in the name of public health corrective actions.
Still, many workplaces suffered, and all were forced to consent to a new normal of remote work. While many American businesses are recovering, these new health and safety protocols are becoming the new standards for employee engagement as a precautionary measure for future sicknesses. Additionally, vaccine rollouts are still progressing internationally, and many people are waiting for their chance to be protected against the Coronavirus.
These transitions were straightforward for those in office buildings and retail, even if they were incredibly taxing on organizational performance. However, for industrial environments around the world, these new mandates have changed the fundamentals of how work is done.
New measures are needed to ensure the safety of these workers for both the sake of their health and the profitability of the operation. Luckily, as many industrial warehouses and factories began to migrate toward automation, they’ve found ways to utilize this technology toward these new priorities, as well as to meet existing goals.
In this article, we’ll discuss how machine vision is enabling factories to ensure the health and safety of their workers both from sickness and from injury through the innovative use of artificial intelligence and machine vision.
What Can Vision AI Do In My Warehouse?
While vision AI can monitor products to determine their quality, it can do largely the exact same thing for the people doing that work itself. AI developer Intenseye created their software solution after visiting dozens of established factories and examining their workflows and safety management software.
While they all had immaculate production pipelines, many did not invest in their health and safety teams, who were merely a handful of people working with traditional paperwork to manage the workflows for thousands of employees.
From there they developed a solution that utilized key features like real-time video analytics to equip these teams with video evidence and information to help enforce safety protocols. With the use of many industrial-grade cameras across the facility, these can capture footage and send it to a central AI fr review, where it can be analyzed to develop heatmaps of where incidents most often occur and calculate the total number of occurrences.
This enables the safety experts team to set Key Performance Indicators and goals so they have a tangible means for measuring if improvements to the safety program are actually made.
They broke the features of their offering down, which would also apply to similar workplace safety software:
Automated Inspections: These turn the cameras that warehouses already utilize into resourceful continuous improvement supervisors with the help of 24/7 automated inspections that are powered by AI.
Real-time compliance tracking: Incident prevention cameras track when employees perform non-compliant behaviors and create numeric scores to represent compliance based on users, tasks, or location.
Simple deployment: This AI system is able to be utilized on existing machine vision cameras, allowing it to be seamlessly integrated into current systems without hassle.
Real-time alert notifications: Instantaneous notifications to email, smart devices, speakers, or even SMS to notify safety teams and workers of potentially serious violations that could cause serious injury.
These software solutions are an incredible feat of software engineering that help promote safety for the sake of workers and warehouses. By utilizing existing vision hardware they can arm safety professionals with unprecedented information and expertise that is both actionable and easily accessible. If that pitch were not compelling, there are several business reasons to consider implementing such a system.
Cost of damages - Facilities can avoid paying to treat damaged property and machinery through repairs or replacements.
Loss in productivity - Injured employees may cause absence time or the need to reschedule, which can reduce your output and the cost efficiency of your factory operation.
Increase in premiums - Warehouses will have to pay increased insurance premiums based on the number of OSHA filings they have made, and they may also have to pay OSHA itself for legal fees.
Loss of goodwill - As recent events have shown, factory incidents can become viral and lead to product boycotts and other negative publicity. By utilizing safety features, these unnecessary incidents can be completely avoided.
The introduction of these systems does indeed require some investment to buy the monitoring equipment, software and train your safety team to use these things. However, that investment will easily be offset in savings from the above-mentioned areas, as it will help to avoid countless costs later.
Can Machine Vision Detect Sickness?
As industrial warehouses resume work in pandemic conditions with masks and social distancing, these are rightfully paired with manual temperature screenings. However, these screens cause increased traffic at factory entrances and exits and are slow affairs that hamper productivity.
Moreso, they fail to detect employees who develop sickness within the warehouse and can still spread it to others around them.
In speaking to a computer vision developer in Australia named Bigmate, Venture Beat’s The Machine reports on how an innovative new solution can remedy this situation. Their solution, appropriately named Thermy, utilizes real-time thermal imaging that can scan 30 people a second to detect body temperatures. These scans are updated 8.3 times every second to ensure accuracy and capture more than just the temperature of the skin.
Bigmate explains that their platform can calculate a body’s core temperature by using high-fidelity computer vision lenses with thermal sensors. This process begins by isolating the person’s head to capture skin temperature and then factoring in other potential variables to understand the core body temperature to ensure it is at an acceptable level.
Since before the coronavirus outbreak, the company had been working on this to curb traditional sickness in the workplace. Now the pandemic has offered them an opportunity to rapidly test and iterate on the software, improving it exponentially. These have now become available and can be implemented at any location within an organization, with all information being sent to a cloud destination for remote analysis.
What Kinds of Workplaces Can Vision AI Improve?
While we’ve spoken primarily about warehouse and industrial implementations of vision AI, this technology can be used in many types of workplaces. Almost all public spaces allow persons and vehicles to come into contact with one another, potentially allowing for dangerous situations to occur.
As Compunnel Digital helps explain, machine vision can help identify risk and arm enforcement and safety officials with the necessary information to act appropriately.
While some might think that businesses and restaurants cannot monitor or enforce mask guidelines, this is simply untrue. Machine vision and cameras allow retailers to supervise the distance maintained between employees and public customers in real-time.
This footage can be recorded for local enforcement to compile and identify if certain individuals repeatedly refuse to follow mask mandates, or for employees, this can be done by safety officials in the workplace.
This is an important metric to understand, as businesses will want to improve their social distancing index as a metric to ensure employee safety. As this visual information is acquired and compiled, artificial intelligence is able to extract numerical values for how well the workplace is performing in terms of social distancing and preventing the spread of Covid 19 or other infections.
With this metric, employers can offer reassurance to existing employees and cite it when recruiting skeptical new hires.
Healthcare environments are particularly delicate, and procedures are all appropriately strict about accounting for this. In medical environments, computer vision helps to improve the safety of both patients and staff with the use of thermal sensors to monitor the temperatures of staff to detect oncoming sickness and to ensure that a doctor or nurse practices proper hand hygiene as soon as they enter a patient’s room.
Precision, high-fidelity lenses, and machine vision equipment also serve to aid surgeons during complex surgical procedures in the event of an error. This technology can also alert a nursing staff if a patient is in need of attention for any reason and can allow many rooms to be monitored from a central location.
This does not replace clinicians but instead augments their purpose and helps to make them more efficient in what they do, rather than requiring them to pace halls and make rounds repeatedly.
As the broadest and open-ended space, machine vision has a lot of crucial applications in public to help with general safety. Here, cameras can monitor individuals and spaces to ensure that safe distances are being maintained, masks are being worn, and tracking the spaces themselves.
If people are frequently touching a certain handrail, an AI can advise that it be cleaned, and if objects are moved through spaces, the AI can help track them. With our thermal scanning technology from earlier, we can also identify if any of these individuals are sick themselves and need medical attention or to be sent home.
All of this provides actionable information into the sanitation of our most vulnerable and dangerous spaces, which can affect the most people.
Aside from health purposes, this can also be used to identify unattended packages or suspicious individuals who may be acting erratically. In addition, automated vehicles will make use of public machine vision cameras.
With both the cameras and vehicles connected to Android and iOs mobile devices on the 5G network, these two things can work in tandem to help the vehicle proceed safely through the streets, avoid potential hazards, or avoid pedestrians where needed.
Ways that this technology can be used widely in public are still being discovered and explored, as regulations about how this information should be collected and acted upon remain to be seen.
Manufacturing is a dangerous environment in which vision AI can help keep people healthy and safe from extreme conditions, one another, or dangerous equipment. These facilities are the most controlled, offering the greatest connections to other devices, people, and resources where necessary.
As these devices not only work to help complete essential work tasks safely and more efficiently, they also enable workers to be kept safe by monitoring employee conditions for sickness or unsafe behavior.
If an employee is not armed with their protective equipment, such as a helmet, vest, or mask, safety teams on-site can be alerted to remedy the issue, thus protecting the worker and everyone around them.
While these might seem like minor infractions of protocol, it is important to respect the reason that these procedures exist in the first place and that the employee would rather not learn through experience.
These practices limit liability and danger for the company and the individual, offering mutual benefits to all parties.
With a marriage of hardware and software, these vision systems are able to improve operations in new, innovative ways. Workplaces are able to detect when employees have fevers, when certain surfaces need to be sanitized, or when core workplace tasks are being performed incorrectly. With this information, we can now minimize damage, risk, and delays to workplace productivity to help cut costs and maximize profits.
More implementations for this technology are being discovered regularly, and there are likely many ingenious uses for AI vision that we haven’t even thought of. Perhaps it will enable us to better plan our cities and street layouts or analyze an underwater riverbed ahead of a bridge’s design.