The world is flooded with data about the world thanks to satellites and environmental sensors. But we still have a lot to do, and Fieldstone Bio believes that microbes can change it.
“They have evolved to sense and respond to information. That’s the trillion calculations that are constantly going on around us,” Brandon Fields, co-founder and chief science officer of Fieldstone Bio, told TechCrunch. “How do you take it and actually manipulate it and make a profit for us?”
Fieldstone’s technology came from that question. The startup was founded in 2023 after spinning out of MIT. There, Professor Chris Voigt’s lab developed a method for converting microorganisms into sensors. Scientists have found ways to program the microorganisms to change color when they encounter things of interest, whether nutrients in the soil or mines hidden in the soil, and then detect them.
“The key technique in Chris’ lab is the idea of ’how to actually visualize these cells from afar,'” Fields said.
Fieldstone Bio recently raised $5 million seed funding led by ubiquili ventures from E14 and LDV Capital. Startups are testing their technology in their labs, and funding allows them to test these microorganisms in the real world.
Each strain is tailored to sense certain compounds, such as farm nitrogen and TNT residues from land mines.
“We separate microorganisms from the environment we want to feel,” Fields said. “We just build sensors into DNA pieces and drop them into these different things and see which ones last the longest.”
Once the microorganisms are ready, Fieldstone will broadcast using a drone. After spending time sensing the environment, the company has different drone snapshots of the area, from hours to days depending on the target.
The image is not a normal aerial photograph seen on Google Maps. Rather, they are photographed using what is called a hyperspectral camera. This camera divides visible light and infrared into up to 600 different colors. Fieldstone microbes reflect light at very specific wavelengths, so you can train AI models to search for these signals in data torrents.
“The power of AI comes in because you can use that information to bully these very faint signals and generate some very cool heat maps of microorganisms that detect the environment,” Fields said.
In addition to agricultural and national security applications, Fieldstone also programming microorganisms to detect environmental pollutants such as arsenic, Patrick Stone said.
“Instead of doing core soil samples every 100 feet, then there’s a resolution of 100 feet. You can get a resolution of 1 inch and map exactly where they need to clean things,” he said.
Gene-edited microbial sensors are broadcast on farms, so there’s no doubt that they will raise eyebrows among people opposed to genetic modifications. Fields said the company is in contact with the EPA to ensure that the company complies with regulations.
Fields hopes that over time, the company’s database will be large enough, and that it can train the model to associate other signals in the environment with the data returned by the microorganisms. This allows hyperspectral cameras to detect arsenic contamination without the need to broaden engineered microorganisms, for example.
“In the end, there’s no need to apply any microorganisms at all,” Fields said. “There are drones, planes and satellites, and we are currently gathering information about chemical information on a global scale.”