InSight Instrument Deployment Arm

Overview

The Instrument Deployment Arm placed the InSight lander’s two primary instruments on the surface of Homestead hollow, a site chosen over three years from an initial 16 candidate ellipses in western Elysium Planitia for its flatness and low rock abundance [4], and later georeferenced to 4.502384N, 135.623447E from post-landing HiRISE imagery [5]. It then spent the rest of the mission doing work it was not built for: pressing on regolith, pinning and hammering a stuck penetrator, and throwing soil across the solar arrays to let the wind clean them. It began as the Mars Surveyor 2001 lander arm, a four-degree-of-freedom backhoe with graphite-epoxy links, a scoop-and-grapple end effector, and no force-torque sensor of any kind [2]. That grapple was designed to tolerate about 7 mm of arm positioning error when lifting the Marie Curie rover off the lander deck, deploying it in 3 cm steps until the crowfoot disengaged [2]. The version refurbished to fly on InSight replaced the links with titanium, raised the joint torques, and kept the predecessor’s blind spot: end-effector force still has to be inferred rather than measured. Placement itself is a flight-software problem as much as a mechanical one: the onboard controller computes the gravity-induced tool-frame deflection from the arm’s own link stiffness and any dangling payload mass, then commands the inverse transform so gravity pulls the tool back to the intended pose, which is what let the same code run unchanged between Earth testbeds and Mars [1].

Specifications

Parameter	Value
Degrees of freedom	4: shoulder azimuth, shoulder elevation, elbow, wrist
Reach	1.8 m
Link material	titanium
Payload capacity	9 kg at 1.65 m extension in Mars gravity, equivalent to 33 N
Peak joint torque	35, 120, 65 and 10.5 N·m, joints 1 to 4
End effector	scoop with a 250 cm³ chamber, plus grapple
Scoop chamber	approximated as a 7.5 × 5 × 4.5 cm box plus a half box; bottom blade 7.5 cm wide
Qualified temperature range	-110 to +70 C structure and joints; -60 to +70 C motors

Source: [6].

Every row in that table is a design specification or a design capability, not a measured performance, and the same is true of the 1 cm positioning accuracy and 1.5 cm instrument position knowledge quoted for the deployment system: both are requirements it had to meet, not a demonstrated repeatability [6]. The one placement actually measured against them, the wind and thermal shield stacked over the seismometer, landed within 1.5 cm of its target, a figure obtained only after bundle adjustment of the arm-mounted camera’s workspace mosaics cut the terrain mismatch between overlapping frames from an average of 5.0 mm down to 1.9 mm [18].

Imaging and localization behind the numbers

Every stereo and localization product described above comes from the same pair of cameras: an arm-mounted Instrument Deployment Camera with a 45 by 45 degree field of view and a body-mounted Instrument Context Camera with a 124 by 124 degree fisheye, both Mars Science Laboratory flight-spare Hazcam and Navcam units upgraded to color before flight [11]. Ground processing turned their images into the mosaics, terrain meshes and instrument-placement products used to select the seismometer and mole sites and, later, to measure the imprints and dumps described below [12]. Commanding the arm against those products relied on the same stereo and rendered-view visualization line JPL had built up since Sojourner, adapted rather than reinvented for a lander with no rover eyes of its own [7].

Running the motors below their qualification floor

From sol 550 the joint heaters were not run at night and the motors dropped below -90 C against a -60 C qualification floor, with no degradation and no operational impediment observed through the rest of the arm’s 1,442-sol mission, 30 C of margin below a ground qualification limit established by flying it rather than by testing it, on a single vehicle over one mission rather than as a bound on the design [6].

Force without a force sensor at the wrist

No contact force the InSight arm applied on the surface was measured directly. Every figure comes from reproducing the flight arm pose in a JPL testbed and reading a three-axis load cell in Earth gravity, and the reconstruction shows the force is strongly pose dependent: scoop pushes at the same site gave median forces that varied by a factor of three between a flat pose and a vertical-forearm pose [6]. A single contact-force figure for the arm does not exist, and the 80 N average downward force often quoted as the end effector’s capability is a nominal figure carried over unchanged from the predecessor arm’s own design description [2], not a per-activity measurement.

Cleaning the solar arrays with the scoop

Seven times the arm dumped soil from its scoop onto the lander deck so wind would carry it across the solar arrays and abrade the dust off them, and six of the seven produced a measurable power gain totaling about 15 percent of array output in that period and extending the mission by roughly 242 sols, while the seventh, dumped in low wind, produced none, since the technique depends on wind the operator does not control and the gain is not attributable dump by dump [6].

Assisting the stuck mole

The arm spent nearly two Mars years, sols 87 through 754, running a JPL testbed campaign to diagnose why the HP3 mole would not go deeper and to try pinning, back-cap pushing and scoop pressing to help it [15]. The mole itself was a self-hammering penetrator whose recoil-control design assumed hull friction from loose regolith [10]; before landing, its hammer strokes had been proposed as a seismic source SEIS could use to map subsurface layering [9], and after landing the same strokes instead became the evidence that a cohesive duricrust, not loose regolith, was absorbing none of that recoil and stopping the mole cold [16]. A catalog of major anomalies across every modern NASA Mars surface mission places multi-year, never-fully-resolved cases like this one at the far tail of the cost distribution the catalog otherwise measures in lost sols [8]. That mole recovery effort, and its eventual limits, are covered in full on the InSight HP3 mole page.

Soil properties inferred from arm contact

The arm’s contact record is the input to a soil model built by the HP3 mole team rather than a set of soil measurements in its own right. Duricrust cohesion at the landing site is 2 to 15 kPa from mole penetration resistance and cone penetration theory, corroborated by an independent estimate of about 6 kPa from the arm chopping the crust with its scoop blade [17]. None of that was a surprise in kind, only in degree: pre-landing laboratory tests on Mars regolith simulants had already bracketed friction angle and seismic velocity ranges meant to bound whatever InSight found [3], and a separate pre-landing compilation synthesizing orbital, lander and rover data had forecast a cohesive near-surface layer over looser sand beneath it [13], so the arm-derived cohesion above landed inside those brackets rather than outside them.

Bulk density runs from about 1200 kg/m³ at the surface to 1600 kg/m³ in the deepest sand and gravel reached, a derived model of only the first 40 cm, and penetration resistance in that deepest layer was 5.3 MPa, seven to eighteen times higher than in the overlying duricrust [17]. That same duricrust, struck repeatedly by the mole’s own hammer, gave up its elastic properties to the seismometer listening to the blows: P-wave and S-wave velocities of 114 and 60 m/s, and shear, bulk and Young’s moduli of 4.32, 9.84 and 11.30 MPa [17].

One deliberate mechanical experiment was run on the regolith itself: on sols 1074 to 1075 the arm pushed to four sequential commanded depths from a 1 cm standoff, with current and torque limits set for maximum downward force, and the imprint depth was read from digital elevation models built from IDC images [6].

Where this arm sits among space manipulators

NASA’s most recent catalog of in-space servicing, assembly and manufacturing systems lists contemporary free-flying and station-mounted manipulators built for on-orbit servicing rather than surface deployment [14], a reminder that the IDA’s backhoe geometry and one-shot placement task are a different design problem from the multi-axis servicing arms that catalog otherwise covers.