Technology Marches On

E

entropychaser

Joined
Jun 27, 2017
Messages
1,340
A Swedish company Bioservo now sells a robotic glove called Ironhand. This glove boosts the grip strength of a 30-year-old man by a factor of five. It activates by the "intention" motion of the wearer. Power comes from a back or hip pack. Already used by General Motors, it sells for $6500. Remember, the Paris Fire Brigade used a robot for interior firefighting at Notre Dame Cathedral. The U.S. Army is developing robots that can climb hills and stairs.
 
Machine learning is the use of computer algorithms that improve automatically with experience. Artificial intelligence is a type of machine learning based on the workings of the human brain in areas such as decision making. Hershey already uses ML to save hundreds of thousands of dollars daily by optimizing candy production. Hospital ICU's improve the prediction of septic shock using ML.
Fire department use of ML in dispatching, EMS protocols, and fireground operations to improve efficiency and save money seems obvious.
 
This might explain a lot. You know how some guys on the job are known for being able to take a lot of smoke? Turns out the same can be said of the lowly amoeba. Some can really take a feed.
Researchers in Pittsburgh and Baltimore tested 35,000 amoebas who had been genetically modified to overproduce different proteins. Next, they exposed them all to cigarette smoke. They then identified the amoebas that did best with the smoke and which proteins they produced. One protein named ANT2 was subsequently found to be deficient in smokers, mice exposed to smoke, and people with chronic lung disease. ANT2 is important to cellular metabolism and increases local hydration which allow the lungs to clear mucus and debris.
The obvious next step is to create drugs or gene therapy that can produce ANT2 and other helpful proteins that can reverse or prevent smoke produced lung disease.
 
San Diego based company ModalAI sells a 1.1 lb microdrone. It can fly for 27 minutes autonomously with an engine/computer the size of a credit card. It is designed to operate in GPS-denied locations such as mines, tunnels and similar tight spaces. You can have one for as little as $2,999.
 
California Institute of Technology engineers have developed "e-skin" that can be placed on a robotic ,or human, arm. This low cost additive manufactured sensor could originally detect temperature and pressure. With special nanomaterial that can be printed by inkjet onto a "glove". It can identify dry and liquid-phase substances including explosives, pesticides, nerve agents, and viruses. In testing, within four minutes it could detect standard chemical explosives, nerve agents and a Covid-19 virus protein.

The system-labeled M-Bot- would seem to be of immense utility to first responders.
 
The California Department of Forestry and Fire Protection (Cal Fire) consumes an enormous amount of data to get their job done. I've noted before that they use aerial infrared/GPS mapping of fires, This information is then fed into a supercomputer at University of California- San Diego to produce 6/12/24 hour predictions of future fire behavior.

On the other side of the equation is technology for early detection of fires. In 2018 the US Department of Defense found that they knew about the fatal Camp fire before Cal Fire did- courtesy of spy satellites. This led to a program named Fireguard where the DOD provides Cal Fire with "sanitized" (Cal Fire doesn't need to know if your car registration sticker is up to date) surveillance data from top secret sources ( satellites, aircraft, drones).

Cal Fire has maintained more than a 1,000 cameras on mountaintops for years. These cameras are monitored 24/7 by 21 command centers. In June Cal Fire began a pilot program using AI to monitor the cameras. Human supervisors are training it to learn false positives (fog, dust, prescribed burns). So far, 40% of AI hits have preceded 911 calls. About a dozen valid AI hits never generated any 911 calls.

You can see the cameras on www.alertcalifornia.org.

Cal Fire's goal is to control 95% of all fires in less than 10 acres.
 
Security cameras, transit cameras, Ring cameras; surveillance cameras are everywhere. As noted above, they cover the entire State of California to look for wildfires. The big problem is that we don't have people to look at them 24/7.

That's about to change. The company Pano AI, located in San Francisco, has developed technology that uses artificial intelligence to detect a fire based on TV inputs. The AI program learns to correctly figure out if smoke during the day, or heat wave lengths at night represent an actual fire; and then automatically notifies the appropriate fire authority. This, of course, is simply another type of automatic alarm that uses another type of sensor. The difference is that this sensor has more complex data to use as opposed to a simple temperature reading or pressure drop across a valve datum. It can also integrate this with other information such as weather, time of day, or GPS. Also, the AI computer program can improve performance by feedback from it's mistakes.

Pano AI is already used in 10 States, Australia, British Columbia and surveys 30 million acres. Customers include utilities, ski resorts, and governments.

It seems to me that this technology can be easily scaled to include every camera connected to the internet.
 
You must log in or register to reply here.
Back
Top