TERRAIN GEOMETRY RECONSTRUCTION FROM IMAGERY Experiment Description ---------------------- Lander and rover images of soils, rocks and other surface features were obtained and rover positions logged during traverses. Analysis performed after the mission shall determine terrain feature classes as well as statistical size and location distributions. Data Collected: --------------- The rover logs positions regularly during traverses. The rover took "drive's end" images from each front camera, providing a stereo image of the terrain in front of the rover after every traverse. The lander took a 360 degree panorama of the landing site during the first few days of the mission. In addition, at the end of each sol's rover traverse, the lander camera imaged the rover in the terrain. The telemetry collected by the rover during traverses combine with the rover and lander images provide the data set for classifying terrain features. Lander cameras provided the panorama in color and stereo. The lander camera images of the rover in the terrain are black and white, stereo 'patches' of sufficient size to show the rover with an error bound consistent with modeled rover navigation uncertainty. Lander camera resolution is 1mrad/pixel. Camera images are compressed at 6:1 (nominally for each sol's rover image). This represents approximately 0.5Mbit to 1.0Mbit of data, depending on rover distance from the lander. Rover cameras have resolution of about 3mrad/pixel. The "drive's end" rover images are full frame (768x484 pixels, nominally 127.5 x 94.5 degrees) uncompressed. Each image represents about 3Mbit of data. Each image data packet is proceeded by a 'record position' header which assists in correlating the image with the status of the rover in particular the position of the rover in a lander referenced coordinate system. Rover traverse data is listed under the experiment entitled "Dead Reckoning Sensor Performance". Data collected at "drive's end": Image Data Record Position processed rover X position (mm) 32bits (I) lsb=1 mm processed rover Y position (mm) 32bits (I) lsb=1 mm processed rover heading (BAMs) 16bits (U) lsb=0.0055 deg measured X,Y accelerometers 2, 16bits (I) lsb=0.0009765g measured CCD (camera) temperature 16bits (I) lsb=0.381ohm exposure duration (msec) 16bits (U) lsb=1 msec camera ID 8bits (U) 0 = left 1 = right 2 = rear starting row 16bits (U) 0-500 starting column 16bits (U) 0-767 ending row 16bits (U) ending column 16bits (U) Image Data number of pixels per row 16bits (U) either: image data n, 8bits (U) or: image blocks n, 104bits (U) Image data is collected from each of the front cameras. Data is taken as the result of a command. The header of the image data telemetry packet contains the mission time (5bytes) and the unique command identified (2bytes). The 'take image' command is parameterized as follows: Capture image byte 0 (c) = camera ID (0-2) with camera (c) at byte 1-2 (t) = exposure time (1msec increments) exposure (t), byte 3 (a) = APID (bits 0-6; 8, 9, 10, 24, 25 return region from bit 7, compression mode: 0=none, 1=BTC) (r1,c1) to (r2,c2) byte 4-5 (r1) = start row (1-500) with APID (a) byte 6-7 (c1) = start column (0-767) [compressed] byte 8-9 (r2) = end row byte 10-11 (c2) = end column SOIL MECHANICS Experiment Description ---------------------- Analysis of rover telemetry collected during traverses over terrain types shall be used to determine basic Martian soil mechanics parameters needed for future Mars rover design. These parameters are sinkage, rolling resistance and traction. At specified locations, the rover will be commanded to collect motor currents, inclinometer angles, wheel revolutions, temperatures, and other engineering data while driving a single wheel. Data Collected: --------------- At a given soil location, the rover rotates wheels in the soil measuring the torque output by the motors and the depth of wheel sinkage in the soil to determine the soil cohesion, c, and phi, the angle of internal friction. Now where t = shear strength of soil (force/area) F = shear force of soil = wheel thrust (force) c = cohesion (force/area) phi = angle of internal friction p = tire ground pressure (force/area) A = wheel contact area W = Wheel load (force) t = c + p tan phi tA = cA + pA tan phi F = cA + W tan phi As the wheel sinks, area is increased, but the wheel thrust change may not be perceptible if the soil has low cohesion. To attempt to separate these effects and to find c and phi (and compensate for actuator losses) during the experiment: - load the soil with two different wheels, with different loads, AEF = (W1-W2) tan phi - however, increased wheel load also may change c and phi - so image the wheel ruts: - assists in the determination of phi, as phi A' the angle of repose for cohesionless soils - image the rear wheel engaged with soil: - assists in determination of contact area - yields qualitative information about crust - yields qualitative information about soil deformation - assists in determination of sinkage A representative experiment procedure then was: - move to a soil location, record the rover's position - begin soil mechanics data taking - rotate a front wheel 30degrees (= cycle; other wheels locked) - end soil mechanics data taking - record rover status - begin traverse data taking - back up 1/2 wheel revolution (A'200mm) - stop traverse data taking - take front camera image of soil patch (100mm x 100mm patch) - record rover status - start traverse data taking - move forward 1/2 wheel revolution - end traverse data taking - begin soil mechanics data taking [repeat next two steps 23 times] - rotate same front wheel 30degrees (= cycle) - stop (nominally 0.5 sec) [completes 720degrees spin of front wheel] - stop soil mechanics data taking - record rover status - begin traverse data taking - back up 1/2 wheel revolution - stop traverse data taking - take front camera image of soil patch (100mm x 100mm patch) [repeat next steps 2 more times with selected front wheel then 3 more times with other front wheel] - move to next location (assumed same soil type), record position - begin data taking for soil mechanics experiment [repeat next two steps 24 times] - rotate same front wheel 30degrees (= cycle) - stop (nominally 0.5 sec) [completes 720degrees spin of front wheel] - stop soil mechanics data taking - move to next location (assumed same soil type), record the rover's position - begin soil mechanics data taking [repeat next two steps 24 times] - rotate back wheel 30degrees (= cycle; other wheels locked) - stop (nominally 0.5 sec) [completes 720degrees spin of back wheel] - end soil mechanics data taking - record rover status - take image from rear camera of back wheel in the soil - begin traverse data taking - back up 1/2 wheel revolution (A'200mm) - stop traverse data taking - take rear camera image of soil patch (100mm x 100mm patch) made by the rear wheel The rover can rotate a wheel at a maximum (no load) rate of 1.2RPM and at a high working load rate of 0.6RPM. Each 30degrees rotation of a wheel with the 0.5 sec pause represents on average 7sec of data taking. The full rotation of 720degrees per wheel in the experiment represents then 168sec of data taking. This procedure is repeated 7 times (considering all wheels) for a total of 1176sec of data taking in the soil mechanics experiment. The amount of data accumulated during a given experiment is: 96 + 8 x 16 + 2 x 96 + 80 x 10 x 168 = 0.135Mbit including the beginning and ending status data records and packet data records. For 7 experiments then this is 0.944Mbit of data. At the beginning and at the end of soil mechanics experiment (at a minimum), health status data is collected. This represents an additional 688bits of data per check for a total of 1376bits for each of 7 experiments or about 0.01Mbit. Each image from the rover cameras is a patch of 15degrees x 17degrees or a patch of 117x156 pixels (oversized by 50% in both horizontal and vertical) for a total of 146Kbits. At full resolution, 4 images are taken for a total of 0.584Mbit. An additional 232bits of engineering data, registering the image, is acquired with each image for a total of 928bits. The rover acquires an image at the rate of 50Kbps so that these images are acquired in about 11.7sec. The traverses performed during the experiment amount to approximately 10 movements of approximately 20cm (not counting the repositioning of the vehicle to each data taking location). The traverse data set represents 408bits of status at the beginning each traverse and 400bits accumulated each wheel radius of travel. Hence, a total of: 10x{408 + 400x[20/6.5]} A' 0.016Mbit* is accumulated. During these traverses we may assume that the rover traverses over nominal terrain at the essentially no load speed of 1.2RPM or 49cm/min. The 10, 20cm movements then represent approximately 4min of traversing. As an estimate of the time required to perform this experiment we may summarize the above as follows: 720degrees of wheel rotation for 7 experiments: 1176sec 19.6min 10, 20cm traverses during the experiments: 245sec 4min 4 images acquired: 11.7sec 0.2min TOTAL: 1432.7sec 23.9min Rover image data is listed under the experiment entitled "Terrain Geometry Reconstruction and Imagery". Rover traverse data is listed under the experiment entitled "Dead Reckoning Sensor Performance". Health status data is listed under the experiment entitled "Logging/Trending". The data for a soil mechanics experiment is listed below. Data collected during soil mechanics experiment: Engineering Data ------------------------------------------------------------------ Record position: first packet of a cycle data record command code, 8bits (U) soil exp = 12 processed rover X position (mm) 32bits (I) lsb=1 mm processed rover Y position (mm) 32bits (I) lsb=1 mm processed rover heading (BAMs) 16bits (U) lsb=0.0055degrees cycle number 8bits (U) 0-99 Begin data taking step for soil mechanics experiment using front wheel time (short) 2, 8bits (U) lsb=1 sec measured X,Y accelerometers 2, 16bits (I) lsb=0.0009765g error state flags 16bits (U) measured front motor temp. (in test) 16bits (I) lsb=0.381ohm measured front steering pot 16bits (I) lsb=0.08deg Begin data taking step for soil mechanics experiment using rear wheel time (short) 2, 8bits (U) lsb=1 sec measured X,Y accelerometers 2, 16bits (I) lsb=0.0009765g error state flags 16bits (U) measured rear steering pot 16bits (I) lsb=0.08deg Data collected while rolling any wheel during experiment (at 10 samples/sec) time (tiny) 2, 8bits (U) lsb=1 centon measured bogie pots 2, 16bits (I) lsb=0.02deg measured differential pot 16bits (I) lsb=0.02deg motor drive encoder 8bits (U) lsb=1 count measured motor drive current 8bits (U) lsb=1.46mA Stop data taking step for soil mechanics experiment using front wheel time (short) 2, 8bits (U) lsb=1 sec measured X,Y accelerometers 2, 8bits (I) lsb=0.0009765g error state flags 16bits (U) measured front motor temp. (in test) 16bits (I) lsb=0.381ohm measured front steering pot 16bits (I) lsb=0.08deg Stop data taking step for soil mechanics experiment using rear wheel time (short) 2, 8bits (U) lsb=1 sec measured X,Y accelerometers 2, 16bits (I) lsb=0.0009765g error state flags 16bits (I) measured rear steering pot 16bits (I) lsb=0.08deg Data records (remaining packets of a cycle) command code 8bits (U) soil exp = 12 cycle number 8bits (U) 0-99 Data is taken as the result of a command. The header of the soil mechanics data telemetry packet contains the mission time (5bytes) and the unique command identified (2bytes). The 'soil mechanics' command is parameterized as follows: Soil mechanics byte 0 (m) = motor ID (0-5) test on wheel (w) byte 1,2 (p) = steering pot position (ignored at position (p), run for center wheel) (-2048 to 2047) for (n) counts, byte 3,4 (n) = encoder counts per cycle repeat (k) times (1-1000) (up to 5 wheel revs) byte 5 (k) = repeat counts (1-1000) DEAD RECKONING SENSOR PERFORMANCE Experiment Description ---------------------- The rover position errors arise from using the dead reckoning (internal state) sensors (including 3-axis accelerometers and a heading gyro) to control mobility. Position error is also a function of distance over various Martian terrain types. The experiment involved traversing paths, noting vehicle behavior through engineering telemetry, and noting differences in visually perceived position versus that output from the dead-reckoning system. Also, visual sensing of terrain types was correlated with the behavior of the vehicle measured from telemetry and an array of several proximity sensors mounted on the rover. Proximity sensors would observe both the clearance between the rover and the ground and the detailed profile of the terrain. Data Collected: --------------- Each time the rover is commanded to perform a movement a set of elemental move or waypoint commands is given to the rover. At execution of each movement the rover: - records rover status - begin data taking during traverse - traverse During waypoint traverses, the rover : - uses laser stripers to do obstacle detection - performs obstacle avoidance by steering away from (perceived) obstacles in its path - update reference heading to reach waypoint commanded. During traverses arising from move or turn commands the rover: - moves the required distance or angle - update reference heading At end of sol's traverses, - rover takes a "drives end" image : full uncompressed image from each rover front camera. - lander images rover in terrain : image is of sufficient size to capture rover in terrain. (lander image size is commandable) During the primary mission, (nominally) the rover traverses in the vicinity of the lander, achieving the objectives of : - getting to soil type for soil mechanics experiment - getting to rock for APXS - getting image(s) of lander Closed loop paths or other TBD paths specifically designed for collecting data for this experiment were expected to be conducted during extended mission. No special experiments were performed during the mission. Waypoint traverses are decomposed into the execution of a series of waypoint commands. During the execution of a waypoint command, data is taken every wheel radius of travel as measured from an average of the wheel encoders. This corresponds to the rate at which proximity detection is performed using the rover laser stripers. The traverse data set represents 408bits of status at the beginning and end of each traverse and 400bits accumulated each wheel radius of travel If there are 10 waypoint commands per traverse, 408 + 400x[100/6.5] A' 0.006Mbit accumulated data per waypoint. At the beginning and at the end of each waypoint command execution (as a worse case), health status data is collected. An estimate of the time required to conduct this experiment is a function of terrain traversed by the rover. We may bound the traversing time between so called nominal terrain, where the rover wheels are driven at the essentially no load speed of 1.2RPM or 49cm/min, and extreme terrain, where the wheels are driven at the high working load speed of 0.6RPM or 24.5cm/min. Hence for a 10m traverse: Terrain Type Time Bit Rate ------------------------------------------------ nominal terrain 20.4min 4.4 Kbit/min extreme terrain 40.8min 2.2 Kbit/min It should be noted that the above are times are for driving of the vehicle alone. Driving is interrupted at each wheel radius of travel with proximity detection and at each vehicle length of travel with the generation of a 'heartbeat' transmission to the lander. Proximity detection requires approximately 5sec to complete while 'heartbeat' transmissions can be conducted in under 1sec (primarily modem power up time). This adds approximately a 37% overhead factor to the time given for nominal terrain traversing and a 26% overhead factor to the time given for extreme terrain traversing. Rover image data is listed under the experiment entitled "Terrain Geometry Reconstruction and Imagery". Health status data is listed under the experiment entitled "Logging/Trending". The data for a traverse is listed below. Data collected during traverses: -------------------------------- Record Status - Traverse : first data packet command code 8bits (U) go to waypoint=8 and turn=10 processed rover X position (start) 32bits (I) lsb=1 mm processed rover Y position (start) 32bits (I) lsb=1 mm processed rover heading (BAM's) 16bits (U) lsb=0.0055degrees error state flags (start) 16bits (U) accumulated average odometer counts 32bits (U) lsb=1 count (start) contact sensor state (start) 16bits (U) bit pattern measured temperature sensors (start) 13, 8bits (I) measured power supply current 9, 8bits (I) lsb=2.03mA to sensors (start) 5.47mA (sensor dependent) measured power supply voltage 10, 8bits (I) lsb=19.52mV to sensors (start) 78.08mV (sensor dependent) Data taking during traverse : every radius of travel time (tiny) 16bits (U) lsb=1 sec processed rover X position ls word 16bits (I) lsb=1 mm processed rover Y position ls word 16bits (I) lsb=1 mm processed rover heading lsbs 8bits (U) lsb=0.0055degrees measured steering pots 4, 8bits (I) lsb=1.3deg measured linear accelerometers 3, 8bits (I) lsb=0.015625g measured motor temperature 2, 8bits (I) lsb=6.1ohm measured motor encoder counts 6, 8bits (U) lsb=1 count measured bogie pots 2, 16bits (I) lsb=0.02deg measured differential pots 16bits (I) lsb=0.02deg measured motor currents 6, 8bits (I) lsb=1.46mA reading laser spot offsets from 15, 8bits (I) 3 lines by 5 nominal: 6 lines by 5 stripes stripes proximity hazard indicator 8bits (I) 0 - 4 Data taking - Traverse : subsequent data packets command code 8bits (U) go to waypoint =8 and turn= 10 error state flags (final) 16bits (U) Data is taken as the result of 'go to waypoint' and 'turn' commands. The header of the traverse data telemetry packet contains the mission time (5bytes) and the unique command identified (2bytes). The 'go to waypoint', 'turn', 'move' commands are parameterized as follows: Go to waypoint at byte 0-3 (x) = destination X (mm N of lander) (x) (y) within (m) byte 4-7 (y) = destination Y (mm E of lander) minutes byte 8 (m) = time limit in minutes (1 - 255) Turn left (n) BAMs byte 0-1 = relative heading in BAMs (+=rt, -=left) Turn rt (n) BAMs Move forward (n) counts with steering (l) (r), Move backward (n) counts with steering (l) (r) Move forward (n) byte 0-1 = encoder counts (+ = fwd, - = back) counts with byte 2-3 = left front/left rear offset from straight steering (l) (r) byte 4-5 = right front/right rear offset from straight Move backward (n) counts with steering (l) (r) SINKAGE Experiment Description ---------------------- Wheel tracks shall be viewed with the rover camera(s) and lander camera to estimate sinkage. During soil mechanics experiments, after driving a single wheel into the soil, the rover and lander images the track pattern produced by this wheel. The bogey angles before and after the single wheel motion is recorded. Data Collected: --------------- During the soil mechanics experiment an image from the rear camera of the rover and also an image from the lander camera, if possible, is taken showing a rear wheel in soil. Also during portions of a soil mechanics experiment a front wheel driven into soil is imaged by the lander or (after the rover has moved out of the wheel track) imaged by the rover cameras. This is correlated with the other measurements taken during the soil mechanics experiments to determine sinkage. The rear camera is in a position to obtain an image of wheel and track together. During the soil mechanics experiment, additional measurement data includes other images of rover wheel patches and portions of tracks, bogey angle measurements and inertial reference. Rover image data is listed under the experiment entitled "Terrain Geometry Reconstruction and Imagery". The data taken during soil mechanics experiments is listed under the experiment entitled "Soil Mechanics". LOGGING/TRENDING Experiment Description ---------------------- All measurable engineering parameters (drive torques/current, position, voltage, etc) shall be logged and time tagged. Analysis of the logged data shall determine performance, degradation, etc. Data Collected: --------------- During each commanded action, the rover records data. In addition, during each sol (nominally, every 10min during day, every hour during the night) the rover records health status data. This data set, augmented by data accumulated during traverses, contains information which allows monitoring of: Power - solar array power generation - stored battery power usage - performance of power converters (regulation) Thermal - performance of WEB - external vehicle temperature Communication - rover/lander data transfer performance Mobility (during traverses) - traverse: wheel motor performance - steering: steering motor performance Navigation: (during traverses) - location estimation - hazard/obstacle detection Data recorded for the health status check is a function of level. Routine scheduled health checks are at level 0 and amount to 688bits of data. During a typical sol the data accumulated for 10hrs of daytime and 14hrs of nighttime operation is: (10x6 + 14) x 688 = 0.051Mbit. During traverses, health checks of level 2 are conducted at the end of each segment of travel (i.e., at the end of each command) to record odometry and performance data. During a typical sol, we may assume about 4 additional level 2 health checks are commanded with each check amounting to 1944bits of data or 0.007Mbit of data per sol. Rover traverse data is listed under the experiment entitled "Dead Reckoning Sensor Performance". Data recorded during health status checks ----------------------------------------- Record Health Status - Level 0 health check level 8bits (U) 0 - 5 processed rover X position (mm) 32bits (I) lsb=1 mm processed rover Y position (mm) 32bits (I) lsb=1 mm processed rover heading (BAM's) 16bits (U) lsb=0.0055degrees error state flags 16bits (U) (start of health check) error state flags 16bits (U) (end of health check) derived mission phase 8bits (U) time at last startup (long) 32bits (U) lsb=1 sec cause of last startup 8bits (U) 0 - 6 averaged accumulated odometer 32bits (U) lsb=1 count measured linear accelerometers 3, 8bits (I) lsb=0.015625g turn sensor integrator drift 8bits (I) measured steering pots 4, 8bits (I) lsb=1.3deg measured bogie pots 2, 16bits (I) lsb=0.02deg measured differential pot 16bits (I) lsb=0.02deg measured APXS deployment pot 8bits (I) lsb=1.3deg contact sensor state 16bits (U) bit pattern measured temperature sensors 13, 8bits (I) lsb=6.1ohm measured power supply current 9, 8bits (I) lsb=2.03mA to sensors 5.47mA (sensor dependent) measured power supply voltage 10, 8bits (I) lsb=19.52mV to sensors 78.08mV (sensor dependent) transmitted comm frame count 16bits (U) lsb=1 count received comm frame count 16bits (U) lsb=1 count received comm error counts 16bits (U) lsb=1 count battery power used 3, 16bits (U) Add for level 2 and higher A/D reference levels 3, 8bits (U) ground lsb=9.76mV -5V lsb=78.08m +5V lsb=39.08mV individual wheel odometer counts 6, 32bits (U) lsb=1 count detected comm error counts (by 13, 16bits (U) lsb=1 count error type) failure counts for all devices 62, 8bits (U) 0 - 6 Minimum during last traverse linear accelerometers 3, 8bits (I) lsb=0.015625g bogey pots 2, 16bits (I) lsb=0.02deg differential pot 16bits (U) lsb=0.02deg front wheel motor temperatures 2, 8bits (U) lsb=6.1ohm Maximum during last traverse linear accelerometers 3, 8bits (I) lsb=0.015625g bogey pots 2, 16bits (I) lsb=0.02deg differential pot 16bits (U) lsb=0.02deg front wheel motor temperatures 2, 8bits (U) lsb=6.1ohm drive/steer motor current 10, 8bits (U) lsb=1.46mA (average during last traverse) drive/steer motor current 10, 8bits (U) lsb=1.46mA (maximum during last traverse) Add for level 3 and higher modem current/transmit 9V 8bits (U) lsb=2.03mA APXS electronics current (9V) 8bits (U) lsb=2.03mA laser current (direct) 5, 8bits (U) lsb=2.85mA CCD current (+/-12V) 8bits (U) lsb=2.03mA accelerometer current (+/-12V) 8bits (U) lsb=2.03mA gyro current (5V regulator) 8bits (U) lsb=5.47mA MAE electronics current (5V 8bits (U) lsb=5.47mA regulator) LED current (contact/encoder) (5V 8bits (U) lsb=5.47mA regulator) WEB heater (motor/heater bus 8bits (U) lsb=4.88mA current) motor heaters (each set, 4, 8bits (U) lsb=4.88mA motor/heater bus current) modem heater (motor/heater bus 8bits (U) lsb=4.88mA current) MAE dust cover current (5V 8bits (U) lsb=5.47mA regulator) Data is taken as the result of a health check command. The header of the health check data telemetry packet contains the mission time (5bytes) and the unique command identified (2bytes). The 'health check' command is parameterized as follows: Health check at byte 0 (n) = check level (0 - 4) level (n) THERMAL CHARACTERIZATION Experiment Description ---------------------- Rover thermal behavior as a function of time and operating situation shall be monitored. The data shall be analyzed on the ground to characterize rover thermal behavior. Data Collected: --------------- As part of routine rover health checks, health status is logged, nominally, every 10min during day, and upon command during the night. 13 temperature measurements are collected: SENSOR LOCATION NUMBER External: CCD's 3, one with each CCD Front Wheels 2, one with each motor Solar Panel 1, one at 'watch plate' WEB: Battery Tray 3, one on each string WEB wall 1 Electronics Card 2 Modem 1 Health check occurrence, particularly at night, is programmable. Night health checks are programmed at times giving approximately equal temperature steps of 8degrees on exterior equipment (based on predictions). Rover health check data is listed under the experiment "Logging/Trending" IMAGING SENSOR PERFORMANCE Experiment Description ---------------------- Images obtained from the lander of the rover tracks in the terrain can be correlated with data logs from the rover taken during driving to characterized the rover hazard avoidance capability. Data collected: --------------- Images from rover are taken (nominally) at "drive's end" (at the end of each traverse). Lander images taken once per sol are used in planning the next sol's rover traverse. Given the lander camera resolution and width of tracks on order of 10cm (where the wheel width is 6cm), track images within lander camera view are planned to be taken within approximately 7m of the lander. (At this distance, with the lander cameras about 1.5m above the surface, a wheel patch subtends an image of approximately 3pixels assuming flat terrain.) Lander camera images are compressed at 3:1. These images represent approximately 0.5Mbit to 1.0Mbit of data, depending on rover distance from the lander. Engineering data is collected during traverses which can be correlated to the images of the tracks shown in the terrain. Rover image data is listed under the experiment entitled "Terrain Geometry Reconstruction and Imagery". The data taken during traverse is listed under the experiment entitled "Dead Reckoning Sensor Performance". Rover health check data is listed under the experiment "Logging/Trending". LINK EFFECTIVENESS Experiment Description ---------------------- Using the statistics taken during communication sessions between the rover and lander, determine the effectiveness of the UHF link on Mars. Data Collected: --------------- Command and telemetry transmissions between lander and rover are collections of frames. Additional frames are transferred to establish/maintain transfer protocol. Atypical frame appears as follows: Transfer Frame |--16--|--8--|--8--|------------0, 32, or 2000------------|--16--| | SYNC | FID |FNUM | DATA FIELD | CRC | Error checking during data transfer is conducted at several layers in the communication protocol. The errors that are logged in this experiment are those at the frame layer and are listed below: ERROR CODE DIAGNOSTIC 0 No error 1 Timed out 2 Short frame received 3 CRC error 4 No sync code 5 Bad FID 6 Bad FNUM 7 No command data ready 8 Expected session start 9 Internal software error 10 Data overflow (too many frames) 11 Received abort 12 Modem latchup detected As part of each health check, the following data is transmitted transmitted comm frame count 16bits (U) lsb=1count received comm frame count 16bits (U) lsb=1count received comm error counts 16bits (U) lsb=1count As a part of level 2 and higher health check, the following additional data is transmitted detected comm error counts (by 13, 16bits (U) lsb=1count error type) During primary mission, rover traverses are in the vicinity of the lander (nominally ~10m radius). In addition, lander proposes to collect telemetry TBD on rover/lander communications. ABRASION Experiment Description ---------------------- The abrasive qualities of Martian soil and dust can be derived from the wear observed on strips of material on one rover wheel. Data Collected: --------------- A center wheel of the rover is instrumented with a pattern of material strips and a solar cell. When this wheel is turned, the current from the solar cell is measured. The light impinging on the solar cell is ambient light reflection from the material strips on this wheel. The variation in current from this cell measured during rotation of the center wheel correlates to the amount of material wear and therefore abrasion by the soil. During the mission, a special experiment is conducted to collect a set of measurements from the cell while the center wheel is rotated (other wheels locked). This is similar to the soil mechanics experiment as follows: - move to a location, record the rover's position - begin wheel abrasion data taking - rotate the abraded wheel 720degrees (other wheels locked) 720degrees (or 2 wheel turns) is the default rotation of the abraded wheel per each command. Additional, rotations can be commanded to enhance material wear either by commanding additional wheel abrasion experiments (data is taken) or by commanding the center wheel motor to run (no data is taken). Up to 16 revolutions of the center wheel can be executed by a single 'run motor' command. A health check is taken prior to conducting the wheel abrasion experiment. The rover can rotate a wheel at a maximum (no load) rate of 1.2RPM. The full rotation of 720degrees per wheel in the experiment represents then 100sec of data taking. While the wheel is rotating, 32bits of data is taken every 16 encoder counts. Since the rover encoders are 2000 counts per wheel revolution, for 2 revolutions 250 samples of 32bits is accumulated. The amount of data accumulated during a given experiment is then : 240 + 250 x32 = 0.016Mbit including the rover status data record. If the 'run motor' command is given, a 16bit parameter which is a part of the command allows the center wheel to be rotated forward up to 32000 counts of 16 revolutions. At the 1.2RPM maximum (no load) rate, a full 16 revolutions would take 800sec or 13min to complete. In the early stages of the rover mission to promote abrasion, a 'run motor' command may be commanded. At the beginning of the wheel abrasion experiment, health status data is collected. This represents an additional 688bits of data. Rover traverse data is listed under the experiment entitled " Dead Reckoning Sensor Performance". Health status data is listed under the experiment entitled "Logging/Trending". The data for a wheel abrasion experiment is listed below. Data collected during wheel abrasion experiment: Engineering Data ------------------------------------------------------------------ Record Status command code 8bits (U) wheelabrsion =25 processed rover X position (start) 32bits (I) lsb=1 mm processed rover Y position (start) 32bits (I) lsb=1 mm processed rover heading (BAM's) 16bits (U) lsb=0.0055degrees measured X, Y accelerometers 2, 16bits (I) lsb=0.0009765g measured bogie pots 2, 16bits (I) lsb=0.02deg measured differential pots 16bits (I) lsb=0.02deg WAE sensor gain 8bits (U) 16,32 total odometer count for the 32bits (U) lsb=1 count abraded wheel (start) error state flags (start) 16bits (U) error state flags (end) 16bits (U) Data collected while rolling abraded wheel during experiment (at every 16 encoder counts) measured cell current sensor, 16bits (I) lsb=1.22mV abraded wheel measured motor drive encoder 8bits (U) lsb=1 count measured motor drive current 8bits (I) lsb=1.46mA Data is taken as the result of a command. The header of the wheel abrasion data telemetry packet contains the mission time (5bytes) and the unique command identified (2bytes). The 'wheel abrasion' command is parameterized as follows: Wheel abrasion byte 0 (a) = 0, default 'high gain' setting of 32 with gain (a) 1, 'low gain' setting of 16 When addition abrasion of the material on the center wheel is desired, the 'run motor' command can be used. It is parameterized as follows: Run motor (m) to byte 0 (m) = motor ID (p) byte 1-2 (p) = encoder counts The 'run motor' command generates the 'command acknowledge' telemetry described as follows: command code 8bits (U) run motor=26 error state flags 16bits (U) ADHERENCE Experiment Description ---------------------- The tendency of Martian dust to adhere to rover surfaces, especially solar arrays and detectors, shall be observed. A test solar cell is situated below a dust cover on the solar panel. This cell was routinely measured for current production as part of the monitoring of the solar panel performance. As part of this separate experiment, the cell is measured both with the dust cover opened and closed. The performance degradation of the cell current due to adherence of dust to the cover can be determined. In addition, a quartz crystal monitor nearby was subject to the same dust adherence. A direct measurement of dust accumulation was made. Data Collected: --------------- The experiment was conducted nominally at least once each sol at the end of traverses, each noon, and at other times during the mission. This experiment is conducted after gathering data from a health check. The procedure is described below: - record vehicle heading - power pitch and roll accelerometers and measure vehicle tilt - turn off power - record time - power temperature sensors and measure solar panel temperature at 'watch plate' - power off temperature sensors - measure current from shorted cell (Dust cover assumed to be closed) - measure open solar cell voltage - actuate cover (power for about TBD (nominally 0.5 sec)) - measure current from shorted cell (Dust cover assumed to be opened) - remove power from cover - wait TBD (30sec) for cover to completely close - measure current from shorted cell (Dust cover assumed to be closed) - power-on QCM's - power temperature sensors and measure solar panel temperature at 'watch plate' - wait for 10sec before reading QCM's - read QCM's - power-off QCM's - power off temperature sensors Data collected for this experiment amounts to 264bits. The experiment required approximately 1 minute to be performed. A heath check is performed before the experiment begins and amounts to another 688bits of data collected. Health status data is listed under the experiment entitled "Logging/Trending". The data for the material adherence experiment is listed below. Data recorded for material adherence experiment ----------------------------------------------- command code 8bits (U) mat adh=23 processed rover heading (BAM's) 16bits (U) lsb=0.0055degrees measured X,Y accelerometers 2, 16bits (I) lsb=0.0009765g error state flags (start) 16bits (U) error state flags (final) 16bits (U) measured temperature at watch plate 16bits (I) lsb=0.381ohm (closed cover) measured temperature at watch plate 16bits (I) lsb=0.381ohm (open cover) measured current from shorted cell 16bits (I) lsb=0.15mV (closed cover start) measured current from shorted cell 16bits (I) lsb=0.15mV (open cover) measured current from shorted cell 16bits (I) lsb=0.15mV (closed cover final) measured open cell voltage 16bits (I) lsb=1.22mV QCM reading 16bits (U) Note that shorted cell implemented with ground return to electronics (not common external ground) to increase resolution. Only 12 bits of the 16 bits assigned above are a part of the measurement. The remaining bits to align packets on byte boundaries. Data is taken as the result of a command. The header of the material adherence data telemetry packet contains the mission time (5bytes) and the unique command identified (2bytes). The 'material adherence' command has no parameters. The 'on-time' for the actuation of the dust cover is an adjustable parameter. The default value is 0.5sec, consistent with an 'on-time' for a mid-sol Martian day (0degreesC ambient temperature). This parameter can be adjusted through the issuance of a 'set parameter command. This command and its parameters is described as follows: Set parameter (p) byte 0 (p) = parameter index = (value) byte 1-2 (value) = parameter value ------------------- * The notation [ ] in a calculation means 'greatest integer less than'. Hence [20/6.5] = 3 PAPERS: Soil Mechanics: "Characterization of the Martian Surface Deposits by the Mars Pathfinder Rover, Sojourner", Rover Team, Science, Vol 278, pp: 1765-1768, December 5, 1998. Moore, H.J., Bickler, D., Crisp, J., Eisen, H., Gensler, J., Haldemann, A., Matijevic, J., Pavlics, F., and Reid, L., "Soil-Like Deposits Observed by Sojourner, the Pathfinder Rover", to appear in Science, March 1999. Abrasion Ferguson, D.C., Kolecki, J.C., Siebert, M.W., Wilt, D.M., Matijevic, J.R., "Evidence for Martian Electrostatic Charging and Abrasive Wheel Wear from the Wheel Abrasion Experiment on the Pathfinder Sojourner Rover", to appear in JGR-Planets, February 1999. Adherence Jenkins, P., Landis, G.A., D. Scheiman, Krasowski, M., Oberle. L., and Stevenson, S., "Materials Adherence Experiment: Technology," 32nd Intersociety Energy Conversion Engineering Conference paper IECEC-97339, July 1997. Landis, G.A., Jenkins. P., and Hunter, G., "Materials Adherence Experiment: Early Results," 32nd Intersociety Energy Conversion Engineering Conference paper IECEC-97340, July 1997.