Deep Space Pulsar Navigation System (DSPNS)

Inspiration

The Deep Space Network, a collection of antenna arrays in California, Australia, and Spain, is the only navigation tool for space. It is specific for objects near the Earth, but is highly inaccurate and unreliable deeper in space. A machine learning algorithm that process data received by a smaller EMF detector or device from pulsars could allow for a much more versatile and compact method for navigation in deep space. Pulsars and the pulses of radiation they send show promise as one of the best autonomous forms of navigation data.

What it does

It processes a database of pulsar radiation, similar to the one that a spacecraft would receive, such as the upcoming Orion spacecraft, and process it to find correlations in radiation distance. This allows it to predict the most likely coordinates of the spacecraft.

How I built it

I used the Azure Machine Learning Studio to develop the algorithm pathway, and utilized the ATNF pulsar catalogue.

Challenges I ran into

Training the model was a challenge because of the nature of the dataset and the datapoints (EMF).

Accomplishments that I'm proud of

It is able to interpret radiation data points, translate them into navigation, and could be important in the future of deep space navigation!

What I learned

I learned about how important it is to create autonomous deep space navigation tools that are highly reliable, and that machine learning algorithms can help to create more accurate systems in this regard. By having a system on the spacecraft itself that is actively determining the location, our navigation could extend past what we would have initially thought.

What's next for Deep Space Pulsar Navigation System (DSPNS)

Integration with a light-weight portable device! This is important because every ounce going into space counts. This algorithm is able to stand alone but would be able to be active in space with an EMF detector or other existing device.