4 New & Critical Metrics that Distribution Centers Need for Reopening
Preventing COVID-19 spread with Video Content Analytics
Some manufacturing and distribution centers, especially for e-commerce fulfillment, have been busier than ever in the past three months, due to the increased demand for goods to be shipped directly to consumers who are under stay-at-home restrictions during the COVID-19 pandemic. Other distribution facilities have been closed because they were deemed “non-essential.” Now that some countries and states have relaxed business restrictions, those centers need to prepare for re-opening — carefully. Beyond checking temperatures, supplying personal protective equipment (such as gloves and masks), or facilitating regular cleaning, there are additional requirements including physical distancing, maximum building occupancy rules, mask wearing, and contact tracing, that impact employee safety, health and morale. Workplaces must be able to demonstrate compliance with such governmental guidelines to remain open, so companies must track these metrics and prove their compliance.
When it comes to ensuring facility health and safety, distribution centers must consider behaviors, such as loading or unloading trucks; moving pallets, packages and products from “pick” to “pack;” and interactions with other individuals in the supply chain and order fulfillment process: Delivery drivers, for instance, transport orders directly to homes and businesses, where they are more likely to come into contact with others.
In comparison to public spaces, enclosed work environments offer some advantages when it comes to monitoring behaviors. However, because distribution centers are often very large, with hundreds of employees, supervisors may find it challenging to ensure that employees consistently follow health and safety mandates – even those unrelated to COVID 19, such as wearing hard hats and hi vis vests or avoiding walking in vehicle-only zones. Fortunately, many distribution centers already have extensive video surveillance (CCTV) networks in place and can, therefore, easily implement video content analytics software – a cost-effective way to manage and encourage workplace health and safety, during the COVID pandemic and beyond.
Powered by deep learning and artificial intelligence, video intelligence software extracts, identifies, classifies, and indexes valuable video metadata, so operators can proactively respond to developing situations, productively review footage, and uncover trends based on aggregated video data. Leveraging video intelligence, operators can understand norms and benchmarks, detect anomalies, count people or objects, measure the distance between objects, and, using object classification and recognition, identify people and vehicles. Using a combination of these functions, video analytics operators can make video searchable, actionable, and quantifiable for effective and efficient site management and security.
Physical distancing recommendations are designed to mitigate the risk of COVID 19 infection, by ensuring six feet (2 meters) of separation between people – a challenging requirement to implement and enforce in any workplace. In production and distribution centers, however, these distancing measures, though hard to apply, would also be valuable for preventing work accidents and injuries, even without the considerations of a pandemic.
To identify physical & social distancing violations, managers can use video analysis driven proximity identification to quantify and analyze distance between individuals. In real-time, management can proactively prevent exposure and illness spread by triggering rule-based alerts for prolonged proximity between two individuals. Furthermore, in the long term, managers can gather anonymous statistical data about behavior patterns over time and on-site locations, for making informed decisions about staffing, signage and workflows to keep employees safe from any number of health and safety threats.
People Counting & Occupancy Tracking
Complementary to physical distancing, workplaces are making an effort to discourage crowding, with some countries and states issuing occupancy guidelines regarding the number of employees simultaneously permitted on site. Video content analytics software can be used to detect the number of people in defined areas and even trigger people-counting alerts whenever a pre-configured threshold is exceeded. Beyond reducing crowding and maintaining safe occupancy ratings, people counting analysis can also be used to proactively ensure that frequently-used areas, such as employee break areas and restrooms, get cleaned and sanitized based on usage. In addition to real-time alerts, people counting data can be aggregated over time, reviewed as needed and visualized to uncover patterns, such as where employees tend to congregate, enter and exit, which enables operations teams to utilize space more effectively and plan based on actionable and quantifiable intelligence.
Face Mask Identification
In addition to proximity and count-based alerts, facility managers can proactively prevent the spread of illness by tracking employee compliance regarding wearing protective gear, such as hi vis vests, hard hats and – now especially – face masks. By configuring alerts for whether or not a hat or face mask is detected, managers can more easily and quickly identify violations and intervene. As with proximity identification, this data can also be aggregated and visualized over time, so that managers can analyze dashboards and understand PPE wearing patterns, identify non-compliance hotspots and develop plans to increase security and safety for employees.
Beyond real-time response, intelligent video surveillance is an important tool for understanding workplace incidents and effectively investigating them. This includes, among other applications, contact tracing for when employees self-identify as having COVID-19. In general, workplaces should develop policies and procedures for the efficient identification and isolation of sick and at-risk individuals. They should have mechanisms in place to conduct contact tracing and quell further outbreaks.
With video content analytics, the facility can forensically review video data to identify and trace diagnosed individuals. By searching and filtering based on face recognition or appearance similarity, system operators can track infected employees over the course of the relevant time span, and combine proximity identification to determine where and with whom that employee interacted, and whether he/she put anyone at risk due to close contact or by not wearing a mask. Keeping the anonymity of the infected individual, the facility could notify relevant employees and recommend next steps for maintaining safety and getting access to healthcare, as necessary. For unaffected employees, business can continue as usual, while those who were exposed can protect their family, friends, and co-workers by voluntarily self-quarantining.
Of course, video content analysis isn’t just applicable for managing the spread of COVID. Having extensive forensic and investigative capabilities helps distribution centers, production facilities and warehouses, reduce inventory loss, improve traffic, and improve incident response times by increasing situational awareness. With video analytics, they can maximize their investment in video surveillance, optimize operations and security, as well as improve worker health and safety.
Striking the balance between worker safety and operational efficiency is always important, but especially so when new risks evolve. The Coronavirus outbreak has helped many businesses reevaluate worker safety and operational efficiencies, identify where improvements can be made, and catalyze change. For organizations that need to demonstrate and prove compliance with health and safety regulations, video analytics makes reporting easy and quantitative. Companies would be smart to apply video content analytics, not only to reduce the risk of COVID transmission in their distribution centers, but also to emerge from the pandemic stronger than they were before.