Planetas Cercanos y Lejanos

Tanto en nuestros planetas vecinos, como en aquellos solo visibles a traves de los más potentes telescopios, otros cuerpos planetarios han fascinado desde siempre a científicos, artistas y exploradores. Esta categoría de la competición te retará a que uses datos de la NASA para explorar los sistemas de otros planetas cercanos y lejanos.

Build a Planet Workshop

EL DESAFÍO

Your challenge is to create a game that will allow players to customize the characteristics of a star and design planets that could reasonably exist in that star system. Ensure that this game provides an educational experience for players!

Background

We live in a solar system with eight planets that are extremely diverse in both size and chemistry. We now know that many of our stellar neighbors have planetary systems of their own. These extrasolar planets, or exoplanets, are discovered through a variety of methods, but primarily through two techniques: measuring the changing brightness of a star due to a planetary transit, or measuring the Doppler shift in the star’s light caused by the changing radial velocity of the star due to a planet’s gravity.

Thanks to satellites such as the Kepler Space Telescope, the Spitzer Space Telescope, and the Transiting Exoplanet Survey Satellite, thousands of exoplanets have been detected and characterized, with a huge variety in characteristics. We now know more about relationships among different types of stars and different types of planets than we have ever before. Your challenge is to create a game that will allow players to customize the characteristics of a star and design planets that could reasonably exist in that star system. Some key customizable features could include size, orbit, atmospheric and surface chemistry, and moons. Ensure that this game provides an educational experience for users on where certain types of planets can exist and what parameters may be necessary for a planet that can harbor life.

Potential Considerations

Not all planets are created equal. Realistically, we cannot customize every aspect of a planet, because so many characteristics rely on others. Be sure to include reasonable restrictions on planetary design to ensure that the game is educational in the areas of planetary and atmospheric chemistry based on the planet’s relationship with a star. For example, in order to have liquid water, a planet must have an orbit around its star that puts it at a proper temperature for water to exist as a liquid. On your team’s project page, be sure to describe the physical and chemical relationships that inform your game, and let your imagination run wild!

Out of This World!

EL DESAFÍO

Crea una app para pilotar un sistema aéreo no tripulado (UAS), como un drone espacial de la NASA, utilizando el sensor giroscópico de 6 ejes de un smartphone o una tablet. La aplicación de pilotaje puede ser combinada con múltiples sensores para mejorar la precisión del vuelo y con las mejores técnicas de maniobra para drones en otros planetas.

Background

Engineers are developing drones for off-world planetary exploration. The flight pattern on these drones will be controlled autonomously. But what if something goes wrong during mid-flight and the autonomous controls do not work correctly? Multiple complications could arise. The drone could start to lose altitude, fly off course into dangerous terrain and weather, or have an uncontrolled landing while engineers are developing new solutions for autonomous controls to be repaired. In the meantime, a drone pilot could take over control of the drone for the safety and protection of multi-million- or multi-billion-dollar equipment and experiments that are aboard the drone.

However, there would not be a human on the planet to repair the drone, so imminent response manual control is vital to the mission(s). The transmissions delay between Earth and the exploratory planet will have a huge impact on the response, so intuitive controls will be ideal for the pilot and drone.

Future transmission technologies are becoming more accessible locally, globally and universally. The WiFi transmissions within a smartphone are becoming broader and are being developed for more distinctive uses in the near future.

This challenge encourages the programmer within you to design and implement new algorithms for the gyro sensor within a smartphone or tablet for drone piloting. The gyro sensor can be used in multiple ways for piloting a drone, but a basic layout has yet to be developed. The primary goal is to find a natural feel among the pilot, controller, and drone.

Potential Considerations

You may (but are not required to) consider the following when designing your system:

  • Basic controls for piloting a drone.
    • Throttle up: Ascend (raises the drone’s altitude above a surface)
    • Throttle down: Descend (lowers the drone’s altitude above a surface)
    • Yaw right: Rotate the drone clockwise
    • Yaw left: Rotate the drone counter-clockwise
    • Roll right: Tilt the drone to the right
    • Roll left: Tilt the drone to the left
    • Pitch up: Tilt the front of the drone up (reverse)
    • Pitch down: Tilt the front of the drone down (forward)

 

  • Example of controls A
    • Throttle up: Raise the smartphone
    • Throttle down: Lower the smartphone
    • Yaw right: Rotate the smartphone clockwise
    • Yaw left: Rotate the smartphone counterclockwise
    • Roll right: Tilt the smartphone to the right
    • Roll left: Tilt the smartphone to the left
    • Pitch up: Tilt the top of the smartphone up (reverse)
    • Pitch down: Tilt the top of the smartphone down (forward)
  • Example of controls B
    • Combination of Gyroscope and Voice
      • Throttle up: Say “Throttle plus 50%”
      • Throttle down: Say “Throttle minus 50%”
      • Yaw right: Say “Right yaw plus 10%”
      • Yaw left: Say “Left yaw plus 10%”
      • Roll right: Tilt the smartphone to the right
      • Roll left: Tilt the smartphone to the left
      • Pitch up: Tilt the top of the smartphone up (reverse)
      • Pitch down: Tilt the top of the smartphone down (forward)
  • Example of controls C
    • Combination of Gyroscope and Touch
      • Throttle up: Slide thumb up on virtual joystick
      • Throttle down: Slide thumb down on virtual joystick
      • Yaw right: Rotate the smartphone clockwise
      • Yaw left: Rotate the smartphone counterclockwise
      • Roll right: Slide thumb right on virtual joystick
      • Roll left: Slide thumb left on virtual joystick
      • Pitch up: Tilt the top of the smartphone up (reverse)
      • Pitch down: Tilt the top of the smartphone down (forward)
  • Example of controls D
    • Combination of Gyroscope, Touch, and Voice
      • Throttle up: Slide thumb up on virtual joystick
      • Throttle down: Slide thumb down on virtual joystick
      • Yaw right: Rotate the smartphone clockwise
      • Yaw left: Rotate the smartphone counterclockwise
      • Roll right: Slide thumb right on virtual joystick
      • Roll left: Slide thumb left on virtual joystick
      • Pitch up: Tilt the top of the smartphone up (reverse)
      • Pitch down: Tilt the top of the smartphone down (forward)
  • Reverse loop: Say “Drone loop backwards”
  • Front loop: Say “Drone loop forward”
  • Barrel roll: Say “Barrel”

You may also consider adding additional sensors to the app to determine other hazards of flight exploration. These include, but are not limited to:

The Memory-Maker

EL DESAFÍO

La electrónica tradicional no funciona bien en Venus. La memoria es uno de los mayores retos. El tuyo es desarrollar soluciones mecánicas que cumplan con las tareas que normalmente se llevan a cabo electrónicamente, en el contexto de la exploración espacial.

Background

With its sulfuric acid clouds, temperatures over 450°C, and 92 times the surface pressure of Earth, Venus is one of the most hostile planetary environments in the solar system. Prior missions have only survived hours! But an automaton (or clockwork mechanical robot) could solve this problem. By utilizing high-temperature alloys, the clockwork rover would survive for months, allowing it to collect and return valuable long-term science data from the surface of Venus. To learn more about the automaton rover, see this link: https://www.nasa.gov/feature/automaton-rover-for-extreme-environments-aree/

There are only a few types of electronics that work in Venus’s hot temperature: those based on silicon carbide and gallium nitride. Unfortunately, the state of the art for these systems is a few hundred transistors (basically the processing power of a solar-powered calculator), so they are highly limited in what they can do, and they consume a lot of power.

As such, your challenge is to develop a mechanical approach to operate a rover on Venus. There are two sub-challenges in this challenge:

  1. A mechanical device that stores and records data: For long-term data storage, consider a system that will store information mechanically. Perhaps it looks like a series of electrical switches, which are turned mechanically on or off, or a re-writable phonograph-style record, or a pinscreen, or something else creatively invented by YOU!
  2. A mechanical device that reverses input power: To back up a current rover with an electric motor, simply reverse the voltage charge, and the motor runs backwards. However, when using mechanical power, the system to provide a reverse gear becomes a bit more complex. Consider a solution that allows for reverse using mechanical power, but that uses an elegant, simple approach with a smaller part count.

Potential Considerations

1. For the Memory Storage Device, you may (but are not required to) consider the following parameters in your solution:

  • Memory: An ideal system can store 1MB of digital (on/off) data.
  • Size: An ideal system fits into a box that is 25cm by 100cm by 100cm.
  • Mass: An ideal memory system is less than 25 kg, including packaging.
  • Electronics: Some simple electrical components may be included, such as wires, resistors, and inductors, if beneficial (since even though most electronics do not work, electricity works fine on Venus). If electronics are used, you may consider a maximum voltage difference across the wires of 18 V or less, and a maximum current of 600 mA (in other words, it essentially can be driven by 2x 9V battery).
  • Signal Input: You may consider the voltage source above, a linear motion of 1 N over 1 cm, or a rotary motion with a torque of 0.1 N-m (the inputs are your choice).
  • Signal Output: You may consider an electrical output using the voltage source above, or a linear motion of 0.25 N, or a rotary motion with a torque of 0.05 N-m (choose whichever output works best for your system).

Other approaches are acceptable for inputting and outputting signal, if explained in the design. You may also choose to use a combination of mechanical and electrical energy. The signal for reading, writing, and erasing memory could be mechanical, electrical, or both.

2. For the Reversing Power Input Device, you may (but are not required to) consider the following parameters in your solution:

  • Design a mechanical reverser that reverses mechanical power direction and considers the following:
    • The ideal input shaft can spin at up to 10 RPM, with a torque of up to 4000 N-m
    • When provided a signal, the ideal output shaft should change directions.
    • Signal will be a mechanical force of 50N with a 3 cm displacement (see prior challenge).
    • A low part count and avoiding linear sliding friction would be key.
    • Size and Mass: The ideal system would be contained within a 300 cm cube and weigh less than 50 kg.
  • Inputs: The ideal input shaft operates at speeds of 0.5 to 10 RPM, with a torque between 1000 to 4000 N-m (if less than 1000 N-m, the system does not need to move).
    • Signal input is a linear sliding pin, with a force of 50N and a linear displacement of 3 cm.
  • Outputs: The ideal output shaft either turns in the same direction as the input, or if the signal is activated, turns in the reverse direction.
  • Electronics: Some simple electrical components may be included, including wires, resistors, and inductors, if beneficial.
  • Optional Electrical Inputs: Up to two wires may be used for voltage inputs. The maximum voltage difference across the wires shall be 18 V or less, and a maximum current of 600 mA (i.e. essentially can be driven by 2x 9V battery).
  • Efficiency: The ideal system is above 85% efficient (i.e. if 4000 N-m of torque is input, the output torque of at least 3400 N-m would be transmitted).

Chasers of the Lost Data

EL DESAFÍO

Ayuda a encontrar maneras de mejorar el rendimiento del machine learning y los modelos predictivos rellenando los huecos en los datasets antes de entrenar a u modelo.

Esto implica encontrar métodos para recuperar computacionalmente o aproximar datos que faltan, debido a problemas con un sensor o ruido en una señal que comprometa la recolección de datos experimentales. Este trabajo se inspira en la recopilación de datos durante los procesos de fabricación aditiva (AM) en los que sensores capturan características de la construcción in situ, pero que tiene aplicaciones en muchas áreas de la NASA.

Background

The data are missing…….

Machine learning (ML) and artificial intelligence (AI) together have the potential to reshape how scientists and engineers use experimental data. Among the many valuable implementations of ML/AI, some examples include autonomy research to find previously undetectable patterns, to supplement or validate physics-based modeling, or otherwise to draw conclusions from very large datasets that would take humans months or even years to process.

A fundamental component of ML and data-driven modeling is having a comprehensive dataset, from which numerous, possibly even hundreds of features can be extracted. The model then “learns” through a process called training how to make predictions based on those features. Every ML algorithm requires vast amounts of data, with complex algorithms like neural networks often requiring thousands of records or observations in a dataset to properly train a model.

While there is great potential in using experimental data for ML/AI, one potential drawback is that experimental data is often compromised during the data collection process. Data collection is driven by sensors monitoring some system, and depending on the experimental environment or setup, those sensors can have limitations. Hardware can be unreliable, unmonitored sensors can fail, and signal noise is always a potential liability. One such example is current research into additive manufacturing (AM) process characterization for material science research. More commonly known as 3-D printing, AM is being explored as a cost-effective and efficient method for creating physical components for aeronautics. However, the in-situ sensor data collected during the AM also captures noise, thereby yielding incomplete datasets.

This issue limits the ability to use ML to predict structure characteristics and to model how parts may perform and therefore understand their structural integrity. Inspired by data loss in AM research with the goal of building accurate ML models, your challenge is to identify ways to computationally recover what is lost when datasets have gaps and excessive noise.

Potential Considerations

Researchers investigate a wide variety of scientific and engineering domains, so methods that can be applied to different types of data from different types of sensors are especially useful. Scientists and engineers also need to ensure that the methods they use in their work can be evaluated and validated by others, so understanding how your approach’s performance might be measured is also helpful. Potential ideas of applications could include (but are not limited to) imputation methods, matrix completion, and tensor completion.

Programming beginners are invited to create a method that can approximate missing data from datasets in Comma Separated Value (CSV) format. Intermediate and advanced programmers are invited to create methods to approximate the missing data and evaluate that method by building a ML model and describing the improvement in that model’s performance before and after the data recovery method was applied.

Listed below are some potential (but not required) additional considerations and recommendations. The challenge of handling missing data is something that confounds many researchers, but is especially critical for implementing machine learning approaches, which use data-driven models to make their predictions.

  • Characteristics vary widely among different datasets; however, an approach that can be generalizable to many different applications would be particularly useful.
  • Methods should consider mixed data (categorical and continuous data types).
  • Source code and models should be open and free for reuse by the public.
  • Code in popular open-source ML programming languages, such as Python or R, is useful.
  • Code should include documentation on model parameters chosen, and why.
  • Although test data for method development may include small datasets, consideration should be given to the feasibility of applications of those methods to large datasets (gigabytes or larger).

Example Resources

NASA data sets suitable for exploring missing data elements are provided in the resources. Each of these can be downloaded in CSV format, and each contains a mixture of categorical and continuous variables. Some amount of data is missing in each of these datasets.

Python libraries for imputation and matrix completion algorithms are provided below. These are examples, not meant to be an all-inclusive list of libraries available. Likewise, these are not the only techniques that can be explored for this problem. Participants are not bound to these libraries listed below and are encouraged to seek out the best possible method for developing their solutions.

NASA in no way endorses any non-U.S. Government entity and is not responsible for information contained on non-U.S. Government websites.

Preferencias de Pprivacidad
When you visit our website, it may store information through your browser from specific services, usually in form of cookies. Here you can change your privacy preferences. Please note that blocking some types of cookies may impact your experience on our website and the services we offer.