Unique Datapath Architecture Yields Real-Time Computing

Information Technology and Software
Unique Datapath Architecture Yields Real-Time Computing (MSC-TOPS-121)
Architecture unencumbers microprocessors, maximizing computational speed for software
Overview
Innovators at NASA Johnson Space Center have developed and successfully flight tested a high-performance computing platform, known as the Descent and Landing Computer (DLC), to suit the demands of safe, autonomous, extraterrestrial spacecraft landings for robotic and human exploration missions. Unique to this platform is a datapath architecture that unencumbers microprocessors by isolating them from input and output interruptions, thus staving off latency and maximizing computational speed for the flight software. To safely land, the DLC must process landing-specific sensor data in real-time and relay this information to the primary flight computer for the spacecraft to avoid environmental hazards like craters and boulders. The datapath architecture presented allows for the DLC’s high-speed computational processing to provide this capability. This technology will be critical for safe access to other surface regions of the solar system in which spacecraft missions could not succeed with current landing capabilities. This datapath architecture technology is at a technology readiness level (TRL) 6 (system/subsystem prototype demonstration in a relevant environment) and is now available for your company to license. Please note that NASA does not manufacture products itself for commercial sale.

The Technology
The DLC platform is composed of three key components: a NASA-designed field programmable gate array (FPGA) board, a NASA-designed multiprocessor on-a-chip (MPSoC) board, and a proprietary datapath that links the boards to available inputs and outputs to enable high-bandwidth data collection and processing. The inertial measurement unit (IMU), camera, Navigation Doppler Lidar (NDL), and Hazard Detection Lidar (HDL) navigation sensors (depicted in the diagram below) are connected to the DLC’s FPGA board. The datapath on this board consists of high-speed serial interfaces for each sensor, which accept the sensor data as input and converts the output to an AXI stream format. The sensor streams are multiplexed into an AXI stream which is then formatted for input to a XAUI high speed serial interface. This interface sends the data to the MPSoC Board, where it is converted back from the XAUI format to a combined AXI stream, and demultiplexed back into individual sensor AXI streams. These AXI streams are then inputted into respective DMA interfaces that provide an interface to the DDRAM on the MPSoC board. This architecture enables real-time high-bandwidth data collection and processing by preserving the MPSoC’s full ability. This sensor datapath architecture may have other potential applications in aerospace and defense, transportation (e.g., autonomous driving), medical, research, and automation/control markets where it could serve as a key component in a high-performance computing platform and/or critical embedded system for integrating, processing, and analyzing large volumes of data in real-time.
Taken from within PowerPoint attachment submitted with NTR. Attachment titled "SPLICE DLC Interface Overview" Shown: a diagram of the Descent and Landing Computer’s (DLC) computing platform with datapath architecture
Benefits
  • Allows unencumbering of system microprocessors to prevent disruptions
  • Enables processing large amounts of data in real-time
  • Designed to OpenVPX standard
  • Incorporates commercially available microprocessors
  • Architecture has potential to be implemented separately for less demanding applications
  • Broadly applicable as an embedded system for a variety of applications
  • Successfully flight tested in DLC application presented

Applications
  • Aerospace & Defense: autonomous landing of air and spacecraft
  • Automotive: autonomous driving and driver-assistance systems
  • Manufacturing: manufacturing process automation and optimization; predictive analytics and failure prognostics
  • Medical: systems for medical sensor arrays, equipment, and instrumentation
  • Research: data acquisition and computational analysis
Technology Details

Information Technology and Software
MSC-TOPS-121
MSC-27099-1
A System to Provide Deterministic Flight Software Operation and Maximize Multicore Processing Performance: The Safe and Precise Landing – Integrated Capabilities Evolution (SPLICE) Datapath, IEEE Space Mission Challenges for Information Technology - IEEE Space Computing Conference (IEEE SMC-IT/SCC 2023), July 18, 2023, NTRS Document ID: 20230001114 . https://ntrs.nasa.gov/citations/20230001114.
Similar Results
The Apollo 11 Lunar Module Eagle, in a landing configuration was photographed in lunar orbit from the Command and Service Module Columbia.
eVTOL UAS with Lunar Lander Trajectory
This NASA-developed eVTOL UAS is a purpose-built, electric, reusable aircraft with rotor/propeller thrust only, designed to fly trajectories with high similarity to those flown by lunar landers. The vehicle has the unique capability to transition into wing borne flight to simulate the cross-range, horizontal approaches of lunar landers. During transition to wing borne flight, the initial transition favors a traditional airplane configuration with the propellers in the front and smaller surfaces in the rear, allowing the vehicle to reach high speeds. However, after achieving wing borne flight, the vehicle can transition to wing borne flight in the opposite (canard) direction. During this mode of operation, the vehicle is controllable, and the propellers can be powered or unpowered. This NASA invention also has the capability to decelerate rapidly during the descent phase (also to simulate lunar lander trajectories). Such rapid deceleration will be required to reduce vehicle velocity in order to turn propellers back on without stalling the blades or catching the propeller vortex. The UAS also has the option of using variable pitch blades which can contribute to the overall controllability of the aircraft and reduce the likelihood of stalling the blades during the deceleration phase. In addition to testing EDL sensors and precision landing payloads, NASA’s innovative eVTOL UAS could be used in applications where fast, precise, and stealthy delivery of payloads to specific ground locations is required, including military applications. This concept of operations could entail deploying the UAS from a larger aircraft.
GEDACS, earth view, mountain elevations
Real-Time, High-Resolution Terrain Information in Computing-Constrained Environments
NASA Armstrong collaborated with the U.S. Air Force to develop algorithms that interpret highly encoded large area terrain maps with geographically user-defined error tolerances. A key feature of the software is its ability to locally decode and render DTMs in real time for a high-performance airplane that may need automatic course correction due to unexpected and dynamic events. Armstrong researchers are integrating the algorithms into a Global Elevation Data Adaptive Compression System (GEDACS) software package, which will enable customized maps from a variety of data sources. How It Works The DTM software achieves its high performance encoding and decoding processes using a unique combination of regular and semi-regular geometric tiling for optimal rendering of a requested map. This tiling allows the software to retain important slope information and continuously and accurately represent the terrain. Maps and decoding logic are integrated into an aircraft's existing onboard computing environment and can operate on a mobile device, an EFB, or flight control and avionics computer systems. Users can adjust the DTM encoding routines and error tolerances to suit evolving platform and mission requirements. Maps can be tailored to flight profiles of a wide range of aircraft, including fighter jets, UAVs, and general aviation aircraft. The DTM and GEDACS software enable the encoding of global digital terrain data into a file size small enough to fit onto a tablet or other handheld/mobile device for next-generation ground collision avoidance. With improved digital terrain data, aircraft could attain better performance. The system monitors the ground approach and an aircraft's ability to maneuver by predicting several multidirectional escape trajectories, a feature that will be particularly advantageous to general aviation aircraft. Why It Is Better Conventional DTM encoding techniques used aboard high-performance aircraft typically achieve relatively low encoding process ratios. Also, the computational complexity of the decoding process can be high, making them unsuitable for the real-time constrained computing environments of high-performance aircraft. Implementation costs are also often prohibitive for general aviation aircraft. This software achieves its high encoding process ratio by intelligently interpreting its maps rather than requiring absolute retention of all data. For example, the DTM software notes the perimeter and depth of a mining pit but ignores contours that are irrelevant based on the climb and turn performance of a particular aircraft and therefore does not waste valuable computational resources. Through this type of intelligent processing, the software eliminates the need to maintain absolute retention of all data and achieves a much higher encoding process ratio than conventional terrain-mapping software. The resulting exceptional encoding process allows users to store a larger library of DTMs in one place, enabling comprehensive map coverage at all times. Additionally, the ability to selectively tailor resolution enables high-fidelity sections of terrain data to be incorporated seamlessly into a map.
NASA UAV
Low Weight Flight Controller Design
Increasing demand for smaller UAVs (e.g., sometimes with wingspans on the order of six inches and weighing less than one pound) generated a need for much smaller flight and sensing equipment. NASA Langley's new sensing and flight control system for small UAVs includes both an active flight control board and an avionics sensor board. Together, these compare the status of the UAVs position, heading, and orientation with the pre-programmed data to determine and apply the flight control inputs needed to maintain the desired course. To satisfy the small form-factor system requirements, micro-electro-mechanical systems (MEMS) are used to realize the various flight control sensing devices. MEMS-based devices are commercially available single-chip devices that lend themselves to easy integration onto a circuit board. The system uses less energy than current systems, allowing solar panels planted on the vehicle to generate the systems power. While the lightweight technology was designed for smaller UAVs, the sensors could be distributed throughout larger UAVs, depending on the application.
https://www.flickr.com/photos/gsfc/47435507852/
SpaceCube 3.0 Mini Processor Card
The SpaceCube 3.0 Mini Processor Card represents orders of magnitude increase in performance and capability over typical radiation-hardened processor-based systems and significant advances over the previous generation of SpaceCube technology. The primary processing engine of the card is a radiation-tolerant FPGA. This processor card is very low weight, can fit within the 1U CubeSat form-factor (10cm x 10cm x 10cm), and will be low power. Much of the SpaceCube 2.0 Micro design is incorporated into the SpaceCube 3.0 Mini design. In addition, lessons learned from the SpaceCube 2.0 Mini card are applied. Instead of using a rigid-flex design, the SC3.0 Mini uses a backplane architecture. The processor card plugs into a backplane that routes signals to other card slots. In order to meet the numerous high-speed I/O interfaces required by the latest generation science instruments and applications, a high-density backplane connector is needed. The SpaceCube 3.0 Mini uses a high-density connector to plug into the backplane. The FPGA has flash memory attached that is used for storing algorithm and application code for any hosted soft processors. The processor card also has a nanominiature front-panel connector that adds even more I/O to support instrument interfaces such as Camera Link or SpaceWire. The SpaceCube 3.0 Mini Processor Card features a rad-tolerant FPGA, but the radiation mitigation can be tailored for harsher environments by adding an external rad-hard device that configures and monitors the FPGA over the backplane. The processor card pushes transceiver quantity, routing, and performance for spaceflight. The card is designed to fit in the compact 1U CubeSat form factor. The SpaceCube 3.0 Mini supports scalability by networking multiple processor cards together.
Spacecube in pieces. SpaceCube is a next-generation computer system developed at the Goddard Space Flight Center in Greenbelt, Md. The potentially revolutionary computer system, which provides up to 25 times the processing power of a typical flight processor, will be testing special software techniques that would make the computer more immune to upsets that happen when radioactive particles affect the computer. The SpaceCube was demonstrated during the Hubble Servicing Mission earlier this year.
SpaceCube 3.0 Flight Processor Card
SpaceCube 3.0 features the rad-tolerant multi-core T2080 processor and the rad-tolerant Kintex UltraScale FPGA. The SpaceCube 3.0 Flight Processor Card meets the industry standards in lightweight systems specifications. In addition, the flight processor card can be installed with an expansion card option to allow a tightly-coupled, mission unique card to be installed. The mission unique expansion card can support a variety of capabilities to make SpaceCube 3.0 a powerful instrument processor, including A/D converters, D/A converters, gigabit ethernet, and additional co-processors. Furthermore, the flight processor card is extremely flexible. Algorithms can be implemented in both the Kintex UltraScale FPGA and the T2080 processor. More sequential portions of the algorithm can be implemented quickly and efficiently on the processor, while other algorithms that are more parallel in nature and computation heavy can be accelerated in the FPGA. Using a hybrid system, each can be optimally implemented to take advantage of the features of both. The SpaceCube 3.0 Flight Processor Card design consists mostly of NASA-qualified flight parts and has many features to mitigate radiation effects on the processor system. The processor card can configure the FPGA to scrub configuration memory. In addition, it can monitor the health of the processors, the FPGA, and any coprocessors on the expansion card using watchdog timers. The FPGA uses error detection and multiple redundant copies to mitigate against radiation upsets to the configuration files, which are stored in external non-volatile memories.
Stay up to date, follow NASA's Technology Transfer Program on:
facebook twitter linkedin youtube
Facebook Logo Twitter Logo Linkedin Logo Youtube Logo