- By Peter Boriskin
- Nov 29, 2023
Just a few years ago, the big trends centered around generating sensor data and creating space to store that data. Today, technology trends like deep learning and generative AI are playing a tremendous role in creating insights out of data. We have big data lakes full of rich information, which, when used responsibly, can be mined and evaluated to uncover patterns, issues, and new uses that can lead to the development of innovative and valuable solutions.
Deep Learning & Generative AI
Right now, we’re trying to understand the tools and deep learning models that go along with data mining and tune them for the greater good of a range of different industries, including security and safety. As exciting as the possibilities are, there’s a need to proceed with caution since some new AI technologies are not perfect yet.
Take chatbots like ChatGPT. These intriguing software applications might be useful for jump-starting research and the creation of subject matter content. Caveats to consider carefully, however, include how effectively these apps have been “trained” and how accurate and trustworthy the information they generate is.
This became evident recently when a New York law firm hit a snag while citing several previous court cases to a judge to prove that there was precedent for putting their client’s personal injury lawsuit on the docket. When the other side informed the Court, they could not find several of the cases referenced in the plaintiff’s brief, the judge determined the cited judicial decisions were bogus. It turned out that, unbeknownst to the plaintiff’s attorney, one of his colleagues had used ChatGPT to do his research.
Of course, there are upsides to deep learning and generative AI. Machine learning, for example, is having a significant positive impact on applications for automated and robotic manufacturing.
Machine vision already uses intelligent cameras to help dramatically improve performance and quality, and robots and cameras don’t get tired or bored with repetitive, tedious inspections. That leaves far less room for error and creates new opportunities for talent to shift to more nuanced, rewarding tasks. Through analytics and deep learning, human operators can be alerted instantly when anomalies occur and respond more effectively.
Intelligent surveillance cameras, analytics, and data also feed into security applications where deep learning can help discern whether an intruder is breaching a perimeter or it’s just an animal passing by. This allows personnel to respond appropriately.
Along with clarity about AI sourcing of information, people in our industry are demanding transparency about the materials and processes going into product development and manufacturing. The design, building, and security communities and their customers are more concerned than ever today about sustainability and reducing the impacts on human health. They want to know about product ingredients, recyclability, and how conscientiously products are being made.
As a result, they require transparency documentation, especially when it comes to green building certification projects like LEED, the Living Building Challenge, WELL, and Passive House, to name a few. Environmental Product Declarations (EPDs), Health Product Declarations (HPDs), Declare Labels, and other third-party verification programs provide key sustainability indicators, attributes, and information that specifiers look for, such as embodied carbon values, manufacturing location, recycled content, Red List Free designations, and more.
Along with sustainability is an increased focus on resilience. Climate change is being looked at as a key contributor to the increase in wildfires and the intensity of storms and flooding. There’s a greater need now for FEMA-rated storm shelters, fire- and water-resistive doors, and blast-resistant openings, particularly for buildings like chemical supply houses where an explosion could be possible.
Cybersecurity transparency is also critical. Organizations want to know how security data is going to be protected at rest and in flight, what tools are being used to secure the data, and whether those tools are standards based. It is now common for customers to ask if a product has been penetration tested by a third party for cybersecurity.
Putting the Plethora of Data to Best Use
The trend of integrating access control and other security data into proptech (property technology) and building management systems continues. Using data generated by all the sensors and devices allows us to make better real estate space decisions, improve efficiency and take security beyond security.
A smart office building, for example, can detect when a scheduled meeting in a conference room didn’t happen or ended early. If no one has badged in for a certain period, no motion is sensed in the room, all the lights are still on, and HVAC is running full bore, integrated access control and building management systems can work together to automatically turn off lights, close window shades, and latch doors to improve energy efficiency and security. Conversely, room scheduling data can activate HVAC to set just the right comfort level prior to a meeting.
Access control can also be integrated with employee management systems. Let’s say a forklift operator in a company’s warehouse or on the manufacturing floor has had too many accidents over a certain period. Data about those mishaps could trigger a need for the employee to undergo more safety training and temporarily restrict the person from using forklifts via lock out/tag out access control.
Ultimately, trends are driven by the evolving needs of customers. By seeking to understand the problems to be solved, we can use technology to help our customers achieve their objectives, enhance security, and improve efficiency.
This article originally appeared in the November / December 2023 issue of Security Today.