TCAS Two, the Resolution Advisory That Overrides Every Voice in the Room, and the ACAS X Architecture Being Built for the Sky That Comes Next

A TCAS resolution advisory (RA) carries regulatory and legal authority that overrides any instruction from air traffic control. When the system commands a climb or descent, the correct response is to comply immediately, maneuver, and notify ATC afterward - in that order, without exception. The collision history that created this rule, and the 2002 Überlingen disaster that reinforced it at the cost of 71 lives, explain why that sequence is non-negotiable.

What TCAS Actually Does - and What It Knows That You Don’t

The Traffic Collision Avoidance System became mandatory on large U.S. commercial aircraft in 1993 under FAA Part 91 regulations. The requirement covers any turbine-powered aircraft with more than 30 seats or a maximum certificated takeoff weight above 12,500 pounds. Most of the developed world has equivalent rules.

TCAS operates independently of air traffic control radar. Every equipped aircraft continuously interrogates the transponders of nearby aircraft using Mode S and Mode C signals. When it receives a reply, it calculates range, bearing, and altitude - building a real-time picture of every transponder-equipped aircraft within roughly 14 nautical miles and ±9,900 feet vertically.

The threat assessment goes beyond simple distance. The system computes time to collision for every tracked intruder by projecting trajectories forward and estimating vertical and horizontal closing rates. The protected zone it applies expands with altitude and speed - at higher flight levels, the system builds in more margin because more reaction time is required.

Traffic Advisory vs. Resolution Advisory: What Each Alert Means

When a threat reaches the first threshold, TCAS issues a Traffic Advisory (TA) - the amber alert. This is information: look for traffic at a specific bearing, range, and relative altitude. Something is developing.

When the geometry continues to worsen, the system crosses a second threshold and issues a Resolution Advisory (RA) - a direct command. Climb. Descend. Increase descent. Maintain vertical speed.

The engineering detail that makes the system function: when two TCAS-equipped aircraft are converging and both systems prepare to issue RAs, they communicate via a data link and coordinate their responses in milliseconds. One aircraft gets Climb. The other gets Descend. Those maneuvers are complementary. The two units negotiated the vertical separation solution between themselves before either pilot heard the alert.

This is why the RA is a command and not a suggestion. The system has already synchronized the solution with the other aircraft. If you deviate from it, you are not making an independent judgment - you are breaking a coordinated maneuver without the other crew’s knowledge.

The Accident History That Built the Case for TCAS

TCAS was developed through the 1980s, funded by the FAA with significant technical work from MIT Lincoln Laboratory. It exists because aviation spent decades proving, with lives, that radar, radio, and human eyes alone were not sufficient.

The collision that triggered the creation of the FAA occurred over the Grand Canyon on June 30, 1956. A United Douglas DC-7 and a TWA Lockheed Super Constellation entered the same airspace at 21,000 feet. Neither crew saw the other until impact. 128 people died. The accident prompted a congressional investigation that produced the FAA and dramatically expanded controlled airspace.

The mid-air collisions continued. Over New York City in 1960, a United DC-8 and a TWA Constellation. Over Indiana in 1969, an Allegheny Convair and a Piper Cherokee on an IFR flight plan. And then the collision that broke something in the collective resolve of the industry: Pacific Southwest Airlines over San Diego in September 1978. A Boeing 727 on approach to Lindbergh Field flew into a Cessna 150 that had drifted into the final approach corridor. 144 people died. The Boeing crew had a brief radar return. There were radio calls. It was not enough.

The question every investigation kept circling was the same: why are we relying entirely on the Mark One Eyeball and three-mile radar resolution to keep aircraft separated?

Überlingen: What Happens When a Crew Doesn’t Follow the RA

July 1, 2002. Southern Germany, over Lake Constance. This case study defines why RA compliance is treated as an absolute standard.

A Bashkirian Airlines Tupolev Tu-154 is at flight level 360, southbound from Moscow to Barcelona - 60 passengers and 9 crew. A DHL Boeing 757 cargo freighter is climbing through flight level 350, crossing the Bashkirian aircraft’s track - 2 crew.

The airspace is managed by Swiss air traffic control company Skyguide. The controller on duty is handling two sectors simultaneously. A short-term conflict alert system is offline for maintenance.

He sees the conflict on radar and issues a descent instruction to the Bashkirian aircraft. Clear, direct, professional. Standard procedure.

At essentially the same moment, both aircraft receive TCAS resolution advisories. The DHL 757’s TCAS commands descend. The Bashkirian Tu-154’s TCAS commands climb. Complementary maneuvers. The system is working exactly as designed.

The DHL crew follows their RA and begins descending.

The Bashkirian crew hears their controller’s voice and begins descending. They do not follow their TCAS resolution advisory. Both aircraft are now descending.

The Tu-154 flies into the DHL 757 at flight level 349. 71 people die.

The German Federal Bureau of Aircraft Accident Investigation (BFU) found that the Bashkirian crew prioritized the controller’s voice instruction over the TCAS resolution advisory. They had been trained that ATC is the authority. In a moment of acute stress and compressed decision time, they went with what they had been taught to trust.

The controller was not negligent. He issued the best guidance he had from his picture of the situation. But he did not know the DHL aircraft was also descending. He did not know the TCAS system had already synchronized a vertical separation solution between the two aircraft. The human coordination loop was slower than the machine coordination loop. When they gave contradictory commands, the wrong one was followed.

After Überlingen, ICAO strengthened its collision avoidance standards. Training programs worldwide added explicit RA compliance drills. The rule became a core element of advanced crew resource management training in every major airline: comply immediately, coordinate with ATC afterward, never the reverse.

Why the RA Knows Something You Don’t

The resolution advisory is not based on the same data available to the pilot. It is not telling you something you could figure out from your own instruments with a few more seconds of thought. It is integrating real-time transponder replies from the intruding aircraft, calculating relative vertical velocities, and projecting where both aircraft will be in the next 20 to 30 seconds.

Your eyes, your altimeter, your traffic display - none of them are running the same calculus at the same update rate.

Überlingen is a case study in what happens when a confident crew believes they have a better read on the situation than the box in front of them. They were wrong, and the other aircraft had no way to know the coordinated solution had been abandoned.

What TCAS Cannot See

The system only tracks transponder-equipped aircraft. A glider without electronics, an ultralight operating below the regulatory floor, a balloon - none of these appear in the TCAS picture. At lower altitudes and in uncontrolled airspace, that gap is real and significant.

The system also requires altitude reporting. A basic Mode A transponder provides an identity squawk but no altitude data. For TCAS to assess vertical separation, it needs Mode C replies encoding pressure altitude. Without that, the system can see traffic but cannot build a complete threat picture.

There is also a deeper structural limit. TCAS was designed for a relatively homogeneous traffic environment: airliners and business jets at high altitudes, moving in predictable performance envelopes, with cooperative transponders. The system was not designed for a sky that includes multirotor drones capable of sudden lateral movements, electric air taxis transitioning between vertical climb and forward flight, or unmanned cargo aircraft whose ground-based pilots introduce response latencies measured in hundreds of milliseconds. That is why ACAS X was built.

What ACAS X Is and Why the Architecture Had to Change

The FAA and Eurocontrol (the European air traffic management organization) have been developing ACAS X for over a decade, with MIT Lincoln Laboratory as the primary technical partner. The fundamental departure from TCAS II is a shift from fixed deterministic logic to optimization-based logic generated by dynamic programming.

TCAS II uses a deterministic logic tree - for a given set of threat conditions, the system looks up the appropriate response in a table built by human engineers. That table covers an enormous number of states, on the order of a billion possible combinations. It is comprehensive, but it is fixed. If the world changes in ways the engineers did not anticipate, the logic tree has no mechanism to adapt.

ACAS X uses a Markov decision process - a mathematical framework for making optimal decisions under uncertainty - combined with a lookup table generated not by hand but by an optimization algorithm running over hundreds of millions of simulated flight scenarios. The algorithm’s objective was to simultaneously minimize mid-air collisions and minimize unnecessary maneuvers - what the industry calls nuisance alerts.

Nuisance alert rate is not a minor concern. If TCAS fires too often when no genuine threat exists, crews habituate to the alerts. Alert fatigue is a documented safety risk with real accident history behind it. ACAS X outperforms TCAS II on both the safety metric and the nuisance alert metric in simulation. That is a meaningful result.

The ACAS X Variants: Built for Different Vehicles

Because ACAS X encodes its logic in a computationally generated table, the optimization can be re-run with different input parameters for different vehicle types.

ACAS Xa targets large commercial air transport - the same mission as TCAS II but with improved performance across both collision prevention and nuisance alert reduction.

ACAS Xu is designed for unmanned aircraft systems. The optimization accounts for different performance envelopes, different response latencies in ground-controlled platforms, and different communication architectures. TCAS II cannot simply be installed on a drone - the physics are wrong and the communication assumptions are wrong. ACAS Xu is being built from the ground up for vehicle classes TCAS II never contemplated.

ACAS Xp is a lighter, simpler variant designed for general aviation, where the full TCAS II certification burden is cost-prohibitive.

When ACAS X Will Actually Arrive

The transition timeline is long. Thousands of aircraft have TCAS II certified and installed, and avionics transitions at fleet scale take 10 to 15 years. ACAS Xa standards are being developed through RTCA (the aviation standards development organization that works alongside the FAA), and regulatory mandates are still years out. The technical work and standard-writing are happening now.

For the advanced air mobility world, ACAS X is load-bearing infrastructure. When electric air taxis approach certification, or when the FAA publishes guidance on urban air mobility corridors, the collision avoidance architecture underneath must handle mixed-performance airspace with vehicles operating under vastly different rules and physics. TCAS II solved an 1980s problem. ACAS X is being built for the problem aviation is about to have.

What This Means for GA Pilots Right Now

For pilots flying aircraft equipped with TCAS, the operational discipline is unchanged. The RA is a command. Comply, maneuver, then tell ATC - in that order. The Überlingen lesson does not belong only to airline crews.

For GA pilots flying with an ADS-B traffic display but no TCAS, that traffic picture is a traffic advisory only. There is no resolution advisory logic and no coordination data link to other aircraft. It is information. The maneuvering decision rests entirely with you. That distinction matters when you are hand-flying through busy airspace deciding whether the amber diamond on your display requires action.

The pattern across 70 years of trying to solve the mid-air collision problem is consistent. The most effective interventions were architectural: build the system to make the right call faster than a human can, train the humans to trust it, and get out of the way when it fires. TCAS II proved that principle works. Überlingen proved what happens when it breaks down. ACAS X is the same principle applied to a more complex future.

Key Takeaways

• TCAS resolution advisories are commands, not suggestions - compliance is immediate and mandatory under FAA regulations and ICAO standards
• When two TCAS-equipped aircraft both receive RAs, the systems coordinate complementary maneuvers via data link before either pilot hears the alert - deviating breaks that coordination without the other crew’s knowledge
• The Überlingen collision on July 1, 2002 killed 71 people when a crew followed an ATC descent instruction instead of their TCAS climb RA; both aircraft ended up descending into each other
• TCAS cannot see aircraft without transponders and cannot fully assess aircraft without Mode C altitude reporting - gaps that matter in uncontrolled airspace
• ACAS X replaces TCAS II’s hand-built logic tree with an optimization-generated table, enabling purpose-built variants for drones (ACAS Xu), commercial aviation (ACAS Xa), and general aviation (ACAS Xp)