Why I started an Image Processing Company: The Grenfell Tower Fire
The charred remains of Grenfell Tower stand only a 40-minute Tube ride from Pudding Lane, the starting point of the infamous London Fire of 1666. Three and a half centuries after one of the biggest disasters in the city’s history, London had gone through it all over again, a rush of fire and death that consumed a heavily populated area while leaving onlookers mostly powerless to control the embers. Even with advanced fire safety regulations and a world-class fire department, Grenfell Tower became the latest horrific example of how quickly fire can spill out of control even in the modern world, leaving experts scratching their heads about how to ensure it never happens again. Given the exploding field of image processing and artificial intelligence, however, is it time to completely rethink fire safety from the ground floor up?
While it’s very difficult to find solace against the vivid images of flames and charred remains, the stunning event has turned into a sobering wakeup call for the rest of the world. Even experienced fire fighters were shocked to see how rapidly the flames burst upward into the top levels of the 24-storey high-rise, and the world looked on in horror and astonishment in real-time as crews struggled to slow down the devastation. Investigators would soon come up with a litany of things that went wrong that night, including a breakdown of the aluminum cladding and insulation that failed to live up to the stringent fire code.
Everyday items causing the spark.
What’s possibly the most startling is that the culprit was an everyday appliance, a simple malfunction of a fridge-freezer that sparked seemingly at random. It was only one of hundreds (if not thousands) of electrical appliances that were likely plugged in at the time of the fire within the building. Despite the centuries worth of advancements since that dark night in September of 1666, the sheer number of potential fire sparks that are jammed into the average high-rise ensures the risk of tragic fire remains every bit as real as before. In the weeks following the fire, building operators all over London were scrambling to install new sprinkler systems, backing up some firms for months and causing others to completely stop taking new clients.
The advantages of going autonomous.
Even though standard sprinkler systems are a tried and proven way to control the devastation of a fire, they’re still antiquated solutions in an expanding digital world. If our electronics can increase our likelihood of a fire, the advancements in autonomous systems can also offer dramatic improvements in the way that we monitor and suppress fires. We now have machine learning, convolutional neural networks, deep learning and narrow AI to automate the mundane and dangerous. We can task computer code and hardware with scanning for fires 24/7 and even make determinations between fire that is properly in-use and an actual fire risk, creating a sophisticated level of surveillance that is always hunting for the next problem.
Much more than just being able to identify hazardous situations, however, image processing and autonomous systems now let us quickly determine the source of a fire and attack the base while the fire is still in its infancy. Although having a manual fire extinguisher on hand is standard in many buildings and apartment complexes, they’re also incredibly prone to user error, particularly in moments of outright panic. With emotions running unchecked in situations of terror, an operator who has never used an extinguisher before might not be able to use it at all, let alone utilize it in time to make a difference. Small situations snowball, ultimately showing that even the strictest of fire codes leave dangerous gaps that can be filled in by tragedy. Even advanced sprinkler systems cannot hone in on the source of a fire like a fire extinguisher powered by a convolutional neural network can, which provides the advantages of visually detecting a fire’s source without the unpredictability of human emotion.
The lessons that can be learned from the Grenfell Tower fire also apply to more than just high-rises and enormous apartment complexes. Often lost in the overwhelming visuals from the main blaze of the fire was the fact that 151 homes in the vicinity went up in the blaze, as the fire swept through the entire adjacent block surrounding Grenfell Tower. This points to a completely different type of danger that anyone living in an area with compact housing has to consider. If your building is up-to-date with smoke detectors and standard extinguishers, a major blaze could still be a significant threat that is beyond your personal control. Even sprinklers, widely seen as top options for fire safety, can dramatically increase property loss and in some cases, often don’t activate until a fire has already caused plenty of damage.
The next frontier of fire protection.
No matter how many things can go wrong, only one thing actually has to go right in order to avoid the next unspeakable tragedy. Autonomy in fire extinguishers offers the next wave of fire safety, a landmark breakthrough that combines the best ways to lower the risk of a deadly fire while eliminating the downsides of the most popular methods. Even though we’ve clearly made some unbelievable strides since the olden days of the 17th century, our fire safety protocol is dated, our building codes not even close to being able to fully live up to the incredible task of preventing the next major fire. It may be tempting to think of great urban fires as a problem of yesterday, but Grenfell Tower stands as a vivid reminder of the many holes left to address when it comes to fighting fires in the modern world. The good news: if we can program machines to make our coffee and drive our cars, we also have the capability of using the autonomous revolution to completely overhaul the way we fight fires.
