We Sent the Camera First
Robert Grand · Battalion Chief who still runs calls
Police chased a suspect into an old 48 inch pipe. We did not know if anyone was alive in there. We did not know if the person was hostile. Forty eight inches is not huge, and the pipe was old. We were not going to send a firefighter into that space on a guess. The police had a small robot with a camera that could navigate the pipe. The camera went in. The person was alive. Then we built a rescue plan around that fact and sent one tech rescue member in on air to get them and pull them out.
The robot did not make the rescue. It made the rescue worth the risk.
By Robert Grand | Battalion Chief, Eugene Springfield Fire | 24 Years in the Fire Service
The Only Question That Mattered
That call was not a fire. It was a confined space with a person we could not see and a threat we could not rule out. The temptation on those jobs is to treat uncertainty as a reason to go faster. Somebody might be dying. Somebody might be waiting to fight. The clock feels like the enemy. The honest first question is smaller than that. Is there a person in there worth sending a person after.
A camera can answer that without putting a firefighter on air in a 48 inch tube. That is not a science project. That is a rule we already used. Do not risk a person unless it is for a person. Until the picture said alive, we did not have a rescue. We had a hole in the ground and a story.
Labs are now running coordinated robot teams through simulated fires and calling the success rates impressive. Griffith University and Cyborg Dynamics showed multi robot teams navigating obstacles and extinguishing simulated fires in a controlled trial (Autonomous swarms of AI-powered robots are here to fight fires, https://www.dpaonthenet.net/article/220700/Autonomous-swarms-of-AI-powered-robots-are-here-to-fight-fires.aspx). Korea’s National Fire Research Institute is building the same idea into doctrine language: drones and robots collect the risk first, then the commander gets a picture worth acting on (Korea looks to AI, drones, robots to make firefighting safer, https://www.koreatimes.co.kr/southkorea/society/20260902/korea-looks-to-ai-drones-robots-to-make-firefighting-safer).
That work is real. It is also not the pipe. The pipe was a camera on a small platform that could fit where a person should not go yet. No swarm. No suppression. One question. Alive or not.
I wrote in July about drones that arrive before the first-due and quietly author the size-up (Who Owns the Size-Up When the Drone Sees It First, https://frontlineintelligence.org/posts/who-owns-the-size-up-when-the-drone). This is the other half of that argument. The machine is allowed to go first when the first job is seeing. It is not allowed to become the judgment. On this call the camera did the only job it was asked to do. It told us there was a life. Then people took over.
What Has to Be True Before a Person Goes In
Trust on that job was not a feeling about robots. It was a short list we could stand on.
The platform had to fit the space. A 48 inch old pipe is not a hallway. If the camera cannot turn the bend, you do not have a tool. You have a story about a tool. The operator had to be able to drive it from outside while we held the mouth of the pipe. The picture had to be good enough to say alive, not good enough to write a press release. And once we knew it was a person, we still needed a rescue plan before anyone went in. Air. One tech rescue member. A way to get them back out. The robot did not pull anyone. A firefighter did.
That is the doctrine I would write if someone asked me tonight. Send the remote platform first when the space is too tight, too unknown, or possibly hostile, and when the question you need answered is whether a human life is actually in there. Keep the human out until the picture gives you a reason to put a human in. Then send the person with a plan, not with hope.
The floor already knows this move. Confined space. Collapse. A void we can look into before we commit. Hazmat that needs a camera before it needs a body. The new part is that the industry is about to sell the same idea as firefighting robots that enter the room and work the fire. Some of that will earn a seat. Some of it will arrive as a demo that has never seen an old pipe.
What has to stay ours is the threshold. The machine goes first to answer a question a person should not have to answer with their lungs. The person goes second because a life is confirmed and a plan exists. If you flip that order, you are not being brave. You are guessing in a space that already told you it was a bad guess.
I do not need a swarm to believe that. I already watched a small camera turn a maybe into a rescue. The future is not robots replacing companies. The future is departments that already know when the camera has earned the first move, and when only a firefighter on air can finish the job.
The picture is allowed to go first. The judgment still belongs to the crew that has to crawl in after it.
Robert Grand is a Battalion Chief at Eugene Springfield Fire with 24 years of service. He writes Frontline Intelligence, a newsletter on operational doctrine, technology, and leadership in Fire & EMS.
If this landed, share it with someone in Fire & EMS who needs to hear it.
Subscribe
From the floor, not the vendor booth. Two times a week.