From the department “sounds like a weapon system” and the NASA JPL:
The concept has been in the development of Dynamic Targeting for more than a decade in the Jet Propulsion Laboratory in Southern California. The first of a series of flight tests occurred in mid -July on board a commercial satellite. The goal: to show the potential of dynamic goals that orbiter can improve the soil image by avoiding clouds and, autonomously, can hunt according to specific, short -lived phenomena such as forest fires, volcanic eruptions and rare storms.
In a recently carried out test, NASA showed how artificial intelligence-based technology could help to circle spaceships that provide targeted and valuable scientific data. For the first time, the technology made it possible for a straw observation satellite to look ahead over its orbital path, to process images quickly and to analyze and determine with the onboard AI where an instrument refers. The entire process took less than 90 seconds without human participation.
This graphic shows how JPL’s dynamic goal uses a LookAhead sensor to see what is on the coming path of a satellite. Onboard algorithms process the data of the sensor and identify clouds in order to avoid interested interest in closer observation if the satellite fits over head.
NASA/JPL-CALTECH
“The idea is to behave more like a person: Instead of just seeing data, it will think about what the data should show and how it should react,” says Steve Chien, a technical fellow in AI at JPL and Principal Investigator for the dynamic targeting project. “When a person burns a picture of trees, he understand that it can indicate a forest fire, not just a collection of red and orange pixels. We try to make the spaceship, the ability to say:” This is a fire “and then focus on the fire.”
This first flight test for dynamic targeting has no hunt for certain phenomena such as fires – that will come later. Instead, the point avoided an ubiquitous phenomenon: clouds.
Most scientific instruments for the area of space vehicles look at what is below. For straw observation satellites with optical sensors, clouds can stand in the way of up to two thirds of the time and block the view on the surface. To overcome this, dynamic targeting provides 300 miles (500 kilometers) and has the ability to distinguish between clouds and clear sky. When the scene is clear, the spaceship forms the surface of the overhead. When it is cloudy, the spaceship cancels the imaging activity to store data storage for another goal.
“If you can be smart what you make, just imagine the floor and skip the clouds. In this way you cannot use all of these images that researchers really do not use,” said Ben Smith from JPL, an Associate with the Earth Science Technology office from NASA, which finances the dynamic target work. “This technology will help scientists get a much higher proportion of the usable data.”
The tests take place on Cognisat-6, a nubbey size introduced in March 2024. During working with Uboatica in 2022, Chien’s team carried out tests on board the international space station, which resembles algorithms with a similar dynamics on the same processor types. The results showed that the combination could work for a room -based remote sensation.
Since Cognisat-6 is missing an educator who deals with a view of the front, the spaceship leans by 40 to 50 degrees to show its optical sensor, a camera that sees both visible and infrarious light near infrarots. As soon as the look shead images have been recorded, the advanced algorithm of dynamic targeting analyzes, which was trained for the identification of clouds. Based on this analysis, the dynamic targeting planning software determines where the sensor for cloud-free views refers. In the meantime, the satellite leans back to Nadir (looks directly under the spaceship) and grabs the planned images, whereby only the floor is recorded.
All of this takes place in 60 to 90 seconds, depending on the original Look-HEAD-WINKEL, since the spacecraft accelerates in a low earth orbit at almost 17,000 miles per hour (7.5 kilometers per second).
With the now proven cloud avoidance function, the next test will be for storms and storms-in essential aim at clouds instead of avoiding them. Another test is to search for thermal anomalies such as forest fires and volcanic eruptions. The JPL team developed unique algorithms for each application.
“This initial provision of dynamic targeting is an extremely important step,” said Chien. “The end goal is the operational use of a scientific mission that leads to a very agile instrument to take new measurements.”
There are several visions for how this could happen – possibly even on space vehicles that examine the solar system. In fact, Chien and his JPL colleagues from another project they had worked were inspired: the use of data from the Rosetta-orbiter of the ESA (the European Space Agency) in order to demonstrate the feasibility of autonomous recognition and imaging of Comet 67p/Churyumov-Gerasimenko.
On earth, the adaptation of the dynamic goal for use with radar could enable scientists to examine dangerous extreme winter weather events that are called deep convective ice storms and are too rare and short -lived in order to be closely observed with existing technologies. Specialized algorithms would identify these dense storm formations with the look of a satellite. Then a powerful, focused radar would turn to keep an eye on the ice clouds to “stare” them, while the spacecraft accelerates through overhead and collects a premium of data over six to eight minutes.
Some ideas include the use of dynamic targeting on several spacecraft: The results of the onboard image analysis from a leading satellite could quickly be communicated to a subsequent satellite that could be commissioned to organize specific phenomena. The data could even be transferred to a constellation of dozens of all -round space. Chien leads a test of this concept, which is later referred to this year federated autonomous measurement.
Like this:
Load…
Related
Do you discover more from watts?
Subscribe to the latest posts to your e -mail.
Comments are closed.