Synthetic Vision Technology, the NASA Research Program That Put a Three-Dimensional Terrain Picture on Your PFD, and the CFIT Trap Hidden Inside the Most Useful Tool in Modern Avionics

Synthetic vision technology (SVT) displays a continuous, three-dimensional terrain picture on a primary flight display by fusing GPS position data, attitude and heading reference system (AHRS) output, and an onboard digital terrain database. It is one of the most consequential safety advances in general aviation avionics of the past 25 years. But the very intuitiveness that makes it powerful contains a dangerous misunderstanding waiting to happen.

Why CFIT Drove the Demand for a Better Solution

Controlled flight into terrain (CFIT) was one of the top categories of fatal general aviation accidents through the 1980s and into the 1990s. The pattern was relentlessly consistent: night flight over terrain, pilot distracted or disoriented, no visual reference to the ground, impact at cruise speed with a ridge or hillside. The aircraft was functioning. The pilot was alive. Nothing in the cockpit had warned them.

Terrain awareness and warning systems (TAWS) addressed part of the problem. TAWS compares radio altimeter readings, barometric altitude, airspeed, gear position, and flap configuration to detect unsafe terrain closure rates. But TAWS is inherently reactive - when it fires, the aircraft is already inside the accident chain. The goal of the system is to provide a few seconds. Those seconds matter, but the design acknowledges that failure has already begun.

How NASA Developed the Concept

NASA researchers at Langley Research Center in Virginia and Ames Research Center in California began work on a different approach in the late 1990s, funded through the NASA Aviation Safety Program - a research initiative targeting the accident categories costing the most lives in civil aviation. Their question was specific: instead of alerting a pilot when they were already dangerously close to terrain, what if the pilot had a continuous three-dimensional picture of all the terrain around the aircraft at all times?

The concept required three data sources working in concert. First, civil GPS, which by the late 1990s could position an aircraft to within a few meters with continuous integrity monitoring. Second, solid-state AHRS sensors that replaced gyroscope-based attitude indicators with continuous, high-precision measurements of pitch, bank, and heading. Third, digital terrain databases - elevation models originally developed for military targeting and geographic information systems, adapted into onboard avionics databases that could be updated on a regular cycle and queried in real time.

None of the three pieces was new in isolation. The innovation was integrating all three into a real-time, three-dimensional display in a certificated avionics system.

The Engineering Behind the Picture

The processing challenge was generating a moving three-dimensional terrain render at display frame rates within the power and space constraints of an avionics box. In the early 2000s, that computing capacity was expensive. NASA’s research partnerships with avionics companies worked through those constraints progressively as processing costs fell and terrain database compression improved without sacrificing accuracy.

The rendering logic is straightforward in principle. The GPS receiver provides position to within a few meters, with integrity monitoring running continuously. The AHRS provides pitch, bank, and heading to high precision. The avionics processor queries the terrain database for elevation data surrounding the aircraft’s current position, then projects that data forward from the pilot’s viewpoint - as if looking out the nose of the aircraft. The display updates in real time: bank left and the terrain picture banks with you; pitch up and the synthetic horizon drops.

Terrain coloring follows an altitude-relative scheme: terrain well below the aircraft is green, rising terrain transitions through yellow to amber, and terrain above the aircraft’s current altitude is red. The color logic functions as a continuous peripheral awareness system - no conscious analysis required. Obstacle data - towers, antennas, structures - is maintained in a separate database and rendered as individually positioned symbols in the synthetic picture. When ADS-B In data is available, many implementations also place traffic targets in three-dimensional space on the display, preserving spatial relationships rather than reducing them to a text readout.

From NASA Research to the General Aviation Fleet

Chelton Flight Systems, based in Boise, Idaho, was among the first companies to move SVT from research prototype to certificated product. Their FlightLogic electronic flight information system received FAA certification in the early 2000s and displayed a full-color synthetic terrain picture - blue sky, green terrain below, amber and red terrain rising toward the aircraft’s altitude - as a core PFD feature.

Chelton’s system worked, and the FAA approved it. But it was expensive and rare enough in the fleet that it reached only a small population of pilots. Changing the safety picture at scale required a wider delivery vehicle.

That vehicle was the Garmin G1000.

Garmin integrated synthetic vision as a standard feature in the G1000 beginning around 2005 and 2006. When new Cessna 172s, Piper Arrows, Beechcraft Bonanzas, and Diamond DA42s left the factory with G1000 cockpits, SVT came with them. Flight schools with new aircraft began training students on glass panels with a three-dimensional terrain picture as a normal, expected feature of the PFD. A generation of student pilots developed intuition for synthetic vision from the first day of training.

What a Night Approach Looks Like With SVT Active

Consider an inbound night approach to an unfamiliar mountain airport. The destination sits in a valley with terrain on three sides. The outside world through the windshield is completely black. On the SVT display, the valley is visible, the surrounding ridgelines are rendered, and the elevation of terrain on the departure end of the runway is coded by color. The synthetic runway box appears on the display well before the actual runway becomes visible.

As the approach continues, the surrounding terrain picture provides a continuous reference for where the high ground is. When the runway lights finally appear through the windshield, the aircraft is already correctly positioned - because the SVT display has been showing the airport’s location and the surrounding terrain throughout the approach. NASA simulation research consistently showed that pilots operating with synthetic vision made significantly fewer inadvertent terrain excursions than pilots without it.

The Limitation That the Picture Conceals

SVT has a fundamental limitation that its intuitiveness can cause pilots to forget: the picture is a model, not a view. It is a database-driven rendering of the world as the database records it. The gap between that model and reality is where serious accidents have occurred.

The terrain database is a snapshot of the world at a specific point in time, updated on a defined cycle. Most certificated terrain databases carry excellent accuracy specifications. But towers get built between update cycles. Temporary construction equipment rises near airports. In documented cases, terrain databases have contained errors in specific geographic regions - particularly areas with less comprehensive original survey data. The system renders what the database contains. It cannot render what the database does not contain.

The deeper hazard is psychological, and it has appeared in National Transportation Safety Board (NTSB) accident investigations often enough to be a documented pattern. Pilots who fly regularly with SVT can begin to process the three-dimensional display as equivalent to looking out the windshield. The picture has depth, color, and moves with the aircraft in accurate spatial relationship to the outside world. Over time, some pilots shift from treating SVT as a supplemental situational awareness tool to treating it as their primary terrain reference.

In instrument meteorological conditions, that shift is dangerous. The SVT display is identical whether the aircraft is in clear air or descending through a cloud layer toward a ridgeline. The GPS works. The AHRS works. The database queries normally. The system has no mechanism to detect instrument conditions or the presence of a cloud layer. The picture will show terrain through the clouds with no indication that those clouds exist.

The NTSB has documented accident cases where SVT data was referenced in pilot decision-making preceding CFIT events. In several, pilots described the SVT picture as part of their assessment that a maneuver or descent was safe. The terrain in the database looked clear. The database could not show the cloud layer obscuring that terrain, or that instrument flight rules should have governed the pilot’s decision from the moment they entered IMC.

FAA advisory material is explicit: SVT is a supplemental situational awareness tool. It does not substitute for instrument flight procedures, minimum safe altitudes, or compliance with IFR requirements.

Database Currency Is a Preflight Go/No-Go Item

Database currency management is a practical discipline SVT demands. Navigation database updates are well established in instrument pilot practice. Terrain and obstacle databases are sometimes managed separately within avionics systems and can fall behind if a pilot isn’t actively tracking them.

Before any flight into mountainous terrain or through areas with significant tower infrastructure, confirming that your terrain and obstacle database is current is not an administrative detail - it is part of the go/no-go. A lapsed terrain database is not a minor discrepancy. It is the specific condition under which the display’s rendering and reality can diverge.

Where the Technology Stands Now

SVT is effectively standard on every new glass cockpit installation. The Garmin G1000 NXi, G3000, G5000, and G3X Touch include it as a built-in feature. Dynon’s SkyView brought SVT to the experimental aircraft community at price points that weren’t possible in the Chelton era. Collins Aerospace’s Pro Line Fusion includes high-definition synthetic terrain rendering in large-cabin business aircraft. The technology has diffused across essentially the entire certificated glass cockpit market.

The current development frontier is enhanced flight vision systems (EFVS). Where SVT generates a picture entirely from a terrain database, EFVS adds a real-time forward-facing camera - typically infrared - whose feed overlays on the synthetic terrain picture. The pilot sees an actual sensed image of the outside world combined with the database-driven render: real runway, real obstacles, real terrain features captured by the infrared sensor alongside the synthetic picture.

EFVS carries a regulatory significance that SVT alone cannot. Under current FAA rules, certain certificated EFVS systems allow descent below decision altitude on an instrument approach, substituting the camera view for the unaided visual contact with the runway environment that standard IFR procedures require. That regulatory credit is not available from synthetic vision alone. High-end business aviation is already moving in this direction.

Further out, sensor fusion research aims to combine EFVS camera data, radar altimeter input, lidar ranging, synthetic terrain, and traffic information into a single integrated primary flight display - an engineered synthesis of every sensor the aircraft can carry, optimized for whatever each sensor does best in the current environment. That work remains in development and high-end prototype stages, but the direction is clear.

Key Takeaways

SVT combines GPS, AHRS, and digital terrain database data to render a continuous real-time 3D terrain picture on the PFD - developed from NASA research in the late 1990s specifically to address the CFIT accident category.
Chelton Flight Systems first certificated the technology in the early 2000s; Garmin’s G1000 integration beginning around 2005–2006 delivered it to the broad GA fleet as a standard feature.
SVT is a model, not a view. The display renders what the terrain database contains and has no ability to detect cloud layers, IMC, or obstacles that post-date the last database update.
Terrain and obstacle database currency is a go/no-go preflight item, not a minor administrative task - particularly for mountain flying or routes with significant tower infrastructure.
Enhanced flight vision systems (EFVS) extend SVT with a real-time infrared camera overlay and, in certificated installations, carry FAA authorization for descent below decision altitude - a regulatory credit synthetic vision alone does not provide.