Failure intelligence, not failure trivia
F Failurepedia

Aerospace and Space Exploration

Mars Polar Lander

NASA's Mars Polar Lander was built to set down near the Martian south pole and dig for water ice with a robotic arm, the first attempted landing in that region. It reached Mars on December 3, 1999, after an eleven-month cruise, entered its descent sequence, and was never heard from again. A review board later concluded that vibration from the landing legs swinging into place was misread by the flight software as ground contact, shutting off the descent engines while the lander was still roughly 40 meters above the surface. It hit the ground at high speed. The loss came two months after its sibling mission, Mars Climate Orbiter, and ended NASA's low-cost "faster, better, cheaper" approach to Mars exploration.

Failed launch Failed initiative Moderate
Company
NASA
Started
1999-01-03
Ended
2000-01-17
Altitude at which the descent engines shut off before touchdown
~40 meters
Collapse speed
Sudden
Preventability
High
Lesson transfer
Industry-wide
Last reviewed
2026-08-06

Narrative

The story

The ambition

Mars Polar Lander was built to do something no spacecraft had done before, set down near the edge of Mars' south polar cap and dig into the frozen ground with a robotic arm to look for water ice. It carried the Mars Volatiles and Climate Surveyor instrument suite, a descent imager, a Russian-supplied light-detection-and-ranging instrument, and a microphone. Bundled on the same launch were two small probes called Deep Space 2, designed to slam directly into the Martian surface and test new penetrator technology. The mission was part of NASA's Mars Surveyor '98 program, a pair of spacecraft, an orbiter and this lander, built under the agency's "faster, better, cheaper" philosophy of low-cost, high-frequency planetary missions.

The rise

The lander launched on January 3, 1999, aboard a Delta II rocket from Cape Canaveral. It flew for eleven months toward a landing region called Planum Australe, near 76 degrees south latitude, chosen because it offered a rare chance to sample material at the edge of the polar cap directly. Along the way it passed the mission's companion spacecraft in the same program, Mars Climate Orbiter, which had launched a month earlier and was meant to study the Martian atmosphere and relay data from landers. Both spacecraft were on track, and NASA had built the program to prove that ambitious planetary science did not require the budgets of the previous generation of Mars missions.

The cracks

The first crack came from the sibling ship. In September 1999, Mars Climate Orbiter approached Mars far lower than planned and was destroyed, the result of a navigation software unit mismatch that was never caught before arrival. Mars Polar Lander was already in flight and could not be redesigned or retested in response. It carried its own unexamined flaw: a design in which sensors on the lander's three legs, meant to signal touchdown so the descent engines would cut off, could be triggered early by the mechanical jolt of the legs swinging down and locking into place during the descent sequence, well before the spacecraft actually reached the ground. The flight software had no way to distinguish that spurious signal from a genuine landing.

The collapse

On December 3, 1999, the lander entered the Martian atmosphere, deployed its parachute, and began powered descent toward Planum Australe. Touchdown was expected around 12:14 p.m. Eastern time, with the first confirming signal anticipated roughly 25 minutes later. It never came. NASA's Deep Space Network listened for a signal that did not arrive, and the orbiter Mars Global Surveyor searched the landing area by camera in mid-December without finding any trace of the lander or its parachute. After more than a month of unsuccessful recovery attempts, including a final set of commands sent January 6 to place the spacecraft in a low-power safe mode in case it was still alive but silent, NASA formally ended the search on January 17, 2000. A Failure Review Board convened to determine what happened, working entirely from lab reconstruction and ground testing because the spacecraft had transmitted no telemetry during its actual descent. In March 2000 the board reported its most probable finding, that the vibration from leg deployment had tripped the touchdown sensors, the software had read that as a landed spacecraft, and the descent engines had shut down while the lander was still falling from roughly 40 meters up. At that altitude, cut off from thrust, it would have struck the surface at somewhere near 22 meters per second, far beyond the roughly 2.4 meters per second the lander's legs were built to survive.

The aftermath

Mars Polar Lander and Deep Space 2 were never located, and no physical evidence of either spacecraft has been confirmed. The back-to-back losses of both Mars Surveyor '98 spacecraft within a few months, one from a navigation unit error and one from a sensor misread, forced NASA to reexamine the "faster, better, cheaper" approach that had produced them. The agency's next Mars lander, the twin-rover Mars Exploration Rovers that followed Spirit and Opportunity, and later Phoenix, returned to more conservative engineering margins, more extensive testing, and closer independent review of critical failure modes. NASA did eventually reach Mars' polar region successfully, when the Phoenix lander touched down there in 2008 using lessons drawn in part from the Polar Lander investigation.

The lessons

A single misread signal, generated by a mechanism working exactly as designed, was enough to destroy an eleven-month, hundred-million-dollar mission in the last forty meters of its flight. The touchdown sensors did not fail; they fired when the legs deployed, precisely as the hardware was built to do. The failure was in the software's inability to tell a known, expected transient from the one signal that mattered. It is a reminder that in a one-shot, no-second-chance landing, the danger is rarely a component breaking outright. It is a system behaving correctly in a context nobody tested for. Because the spacecraft sent back no data during its final minutes, investigators could never confirm the cause with direct evidence, only reconstruct the most likely one, a gap that says something about the cost of cutting margin and instrumentation from a mission that has exactly one chance to succeed.

Causal timeline

Failure Anatomy

  1. 1999-01-03

    Launch

    Mars Polar Lander launches from Cape Canaveral aboard a Delta II rocket, carrying the Deep Space 2 microprobes, bound for a landing near the Martian south pole. [1]

  2. 1999-09

    Sibling spacecraft lost

    Mars Climate Orbiter, the companion mission in the same Mars Surveyor '98 program, is destroyed approaching Mars because of an unconverted units error in its navigation software. [2]

    Poor execution
  3. 1999-12-03

    Descent and silence

    Mars Polar Lander enters the Martian atmosphere and begins its powered descent toward Planum Australe. The expected post-landing signal never arrives. [3] [4]

    Technical failure
  4. 2000-01-17

    Search efforts end

    After more than a month of listening for a signal and searching the landing site by orbital camera, NASA sends final commands and formally concludes all attempts to recover the spacecraft. [5]

  5. 2000-03

    Review board finds the probable cause

    A Failure Review Board, working from lab reconstruction rather than descent telemetry, reports that a spurious touchdown signal from the leg-deployment sensors most likely shut off the descent engines roughly 40 meters above the surface. [6]

    Technical failure
  6. Undated

    End of an era

    The back-to-back losses of both Mars Surveyor '98 spacecraft push NASA to move away from the "faster, better, cheaper" approach and toward more conservative testing and review on later Mars missions. [7]

    Incentive failure

Structured analysis

What Went Wrong

Root causes

A landing sensor misread as touchdown. Sensors on the lander's three legs, meant to signal ground contact so the descent engines would shut off, generated a spurious signal when the legs swung down and locked into place during descent. The flight software could not distinguish this transient from an actual landing and cut the engines while the spacecraft was still falling. [6]

Contributing factors

The failure mode was not caught before launch. Ground testing and review before launch did not identify the vulnerability of the software to a spurious touchdown signal generated during leg deployment, a scenario the review board later reproduced in lab testing after the loss. [6]

Built under cost and schedule pressure. Mars Polar Lander was developed inside NASA's low-cost "faster, better, cheaper" Mars Surveyor '98 program, and funding and management shortfalls were cited as contributing factors in the mission's failure. [7]

Immediate trigger

Descent engines shut off roughly 40 meters up. During the December 3, 1999 landing sequence, the flight software registered a false touchdown signal from the leg sensors and commanded the descent engines to shut down while the spacecraft was still airborne, far above survivable altitude. [6]

Visible symptoms

No signal at the expected time. The first confirming radio signal after touchdown was expected around 12:39 p.m. Eastern time on December 3, 1999, and it never arrived, despite repeated listening by NASA's Deep Space Network. [3] [4]

Warning signs

The sibling mission was lost two months earlier. Mars Climate Orbiter, launched in the same Mars Surveyor '98 program, was destroyed approaching Mars in September 1999 because of an uncaught navigation software error, a sign the program's engineering and review discipline had gaps before Polar Lander's own arrival. [2]

Affected groups

Taxpayers

Contested

Disputed points

Interpretations where credible accounts genuinely differ, presented as disputes, not settled facts.

The leg-deployment sensor explanation is NASA's leading finding, not a confirmed one. Mars Polar Lander transmitted no telemetry during its actual descent or after its expected landing, so the Failure Review Board could not observe the failure directly. It reconstructed the scenario from lab testing that reproduced the spurious signal, and reported it as the most probable cause rather than a proven one. [6]

Unresolved

Keep reading

Evidence

Claims & sources

Every numbered marker in the analysis links to the claim it rests on, and each claim to its sources.

  1. [1]

    Mars Polar Lander launched January 3, 1999, aboard a Delta II rocket from Cape Canaveral, carrying the Deep Space 2 microprobes toward a landing near the Martian south pole.

    High Fact Mars Polar Lander Mars Polar Lander / Deep Space 2 Could Not Have Survived: 20 Years Since NASA's Ill-Fated Mars Polar Lander
  2. [2]

    Mars Climate Orbiter, the companion mission in the same Mars Surveyor '98 program, was lost approaching Mars in September 1999 because of an uncaught navigation software error, two months before Mars Polar Lander's own arrival.

    Moderate Fact Could Not Have Survived: 20 Years Since NASA's Ill-Fated Mars Polar Lander
  3. [3]

    On December 3, 1999, Mars Polar Lander entered the Martian atmosphere and began its descent toward Planum Australe near the south pole, with touchdown expected around 12:14 p.m. Eastern time and a confirming signal expected roughly 25 minutes later.

    High Fact Mars Polar Lander Mars Polar Lander / Deep Space 2 Could Not Have Survived: 20 Years Since NASA's Ill-Fated Mars Polar Lander
  4. [4]

    No signal was ever received from Mars Polar Lander after its expected landing time, and a search by Mars Global Surveyor's camera in mid-December 1999 found no trace of the lander or its parachute.

    High Fact NASA Concludes All Attempts to Communicate with Lander Mars Polar Lander / Deep Space 2
  5. [5]

    NASA sent final commands on January 6, 2000 to place the lander in safe mode and formally concluded all attempts to recover the spacecraft on January 17, 2000.

    High Fact NASA Concludes All Attempts to Communicate with Lander Mars Polar Lander / Deep Space 2
  6. [6]

    A Failure Review Board, reporting in March 2000 and working from lab reconstruction rather than descent telemetry, found the most probable cause of loss was a spurious touchdown signal generated by the landing legs deploying, which the flight software read as ground contact and used to shut off the descent engines while the lander was still roughly 40 meters above the surface, likely striking the ground at around 22 meters per second against a survivable 2.4 meters per second.

    Moderate Reported explanation Mars Polar Lander Mars Polar Lander / Deep Space 2 Mars Polar Lander / Deep Space 2 Could Not Have Survived: 20 Years Since NASA's Ill-Fated Mars Polar Lander
  7. [7]

    Mars Polar Lander was developed within NASA's low-cost "faster, better, cheaper" Mars Surveyor '98 program; inadequate funding and management were cited as contributing factors, and the back-to-back losses of both program spacecraft pushed NASA away from that approach on later missions.

    Moderate Reported explanation Could Not Have Survived: 20 Years Since NASA's Ill-Fated Mars Polar Lander Mars Polar Lander

Sources