Network Interface Links Sensor-Web Instruments
- Created: Monday, 01 July 2013
NASA Airborne Science operates a fleet of aircraft in conjunction with orbiting satellites for Earth observations. In 2004, NASA started planning missions to employ constellations of instruments flying on those platforms that would mutually interact and communicate as a network with stations on the ground. These sensor webs would simultaneously collect data from multiple perspectives to better describe hurricanes, polar ice conditions, and other geophysical dynamics. Data from various spectra and locations could be correlated in real time to form detailed composites of events in progress.
NASA aircraft had for decades collected atmospheric and terrestrial information for a wide range of Earth science research conducted by universities, companies, and government agencies such as the National Oceanic and Atmospheric Administration (NOAA). Pilots switched on the instruments, the data were acquired, and the recordings were unloaded and analyzed after the aircraft landed. To standardize instrument connections to the ER-2 high-altitude aircraft, NASA created the Mark I Experimenter Interface Panel (EIP) in the 1990s, which provided power to the payloads and a link to a pilot-operated switch panel. That EIP was largely a harness pass-through with a small printed circuit board that carried low-level relays for the cockpit switch panel. It simplified the engineering required to deploy instruments to the ER-2, and was later installed on the WB-57 aircraft.
However, the system provided no visibility into the instruments’ power consumption or other onboard conditions. Unless a payload provided one, there was no capability for high-bandwidth data transmission from the NASA Airborne Science aircraft. Researchers had no insight about data or instrument operation during flights, except by audio communication with pilots who often had two lamps per instrument to monitor. A science team could pay tens of thousands of dollars for fuel, only to discover after hours of post-flight data processing, that their instruments malfunctioned or an unforeseen target of interest was missed.
Developing Sensor Net
The development of a new “Sensor Net” payload system (Figure 1) began in 2008, when the Global Hawk Unmanned Aerial Vehicle (UAV) was added to the Airborne Science fleet. This in turn motivated development of a fleet-wide standard electrical interface for research instruments. The Airborne Payload C3 System (APCS) in the Global Hawk incorporated prototypes of a new (Mark II) EIP, a NASA Airborne Science Data Acquisition and Transmission unit (NASDAT), and an Ethernet network by which the NASDAT — the network server — would distribute a precise timing reference, navigation data, and avionics data among experiment instruments. Each instrument would report its status through an EIP via the network to the NASDAT, and each EIP would update the NASDAT via the network about the dc power consumption of the instruments, cockpit switch states, and related ship-side housekeeping details, which the NASDAT would store and rebroadcast to the instruments.
The NASDAT would provide limited communication with ground stations over four Iridium satellite telephone modems. The Link Module, a commercial off-the-shelf unit, would provide 500 Gbytes of mass storage, serve as a gateway to high-bandwidth communications to enable researchers on the ground to view data in real time, and would process some data onboard as they were being collected.
The first flights for the APCS were on the GLOPAC mission aboard the Global Hawk in 2010, which included a flight of record-breaking duration for that UAV. The payload included 11 networked instruments, which scientists used to collect data about cloud structures, Asian dust, and stratospheric air masses that had moved from the North Pole. Observations from the Global Hawk were also used to validate atmospheric data from the Aura and CALIPSO environmental monitoring satellites.
Spanning the Fleet
The Mark II EIPs developed for the Global Hawk were a success, but were optimized for that aircraft and did not span the requirements of the entire fleet, which includes such diverse aircraft as the DC-8, P-3B, B200, the ER-2, the WB-57, and others. In particular, because the Mark II EIP was dedicated to a remotely controlled aircraft, it did not provide an interface for manned cockpit switches. Moreover, it did not support such legacy data-communication protocols as ARINC-429, RS-232, RS-422, or Synchro. Though those protocols and their hardware interfaces may seem outdated, the expense and engineering effort required to convert or redesign the many dozens of existing instruments that incorporate them would be major. Consequently, in 2010, the NASA Airborne Sensor Facility began development of the Mark III EIP (Figure 2), a fleet-wide design that accommodates existing instruments that have legacy interfaces, as well as new payloads with 10BASE-T, 100BASE-TX, or 1000BASE-T data interfaces.