Skip to content

Feat/aaronrhim/hmi deploy layouts - #187

Merged
aaronrhim merged 174 commits into
mainfrom
feat/aaronrhim/hmi_deploy_layouts
Aug 7, 2026
Merged

aaronrhim merged 174 commits into
mainfrom
feat/aaronrhim/hmi_deploy_layouts

Conversation

@aaronrhim

@aaronrhim aaronrhim commented Aug 4, 2026

Copy link
Copy Markdown
Contributor
  • Added cameras to hmi
  • Added lighting
  • Added Ptz camera
  • Added cats
  • Added launch files for hmi
  • Added gnss pathing
  • added morse code output
  • added XOR connector output
  • added task descriptions and calc
  • added triple monitor launch

TODO:

  • IK for arm ***
  • finish arm homing ***

BenUBC and others added 30 commits July 26, 2026 12:50
Brings the Science_RD spectrophotometer workflow into the HMI so standard
curves can be built and unknown samples measured on the rover instead of a
laptop shell.

The Python pipeline (curve_gen.py, capture_and_predict.py, fake_data_gen.py)
is vendored unmodified from Science_RD and stays the single source of
numerical truth; the module drives it over QProcess so a capture never
blocks the UI. spectro_paths keeps scripts and data in the repo rather than
~/.config. Also fixes the container Python env, where apt's matplotlib and
scipy are built against the numpy 1.x ABI and fail to import against the
image's numpy 2.x, leaving savgol_filter and curve_fit unusable.

SpectrumView ports PySpectrometer2's graph to Qt in the HMI theme: 10 nm
graticule with 50 nm labels, the trace filled with each wavelength's visible
colour, labelled peaks, hover readout and an assay-wavelength marker. Real
captures are in pixel order and the calibration polynomial need not be
monotonic — Science_RD's Rezasurin data folds twice, so 570 nm is claimed by
three different pixels — so traces are split at their turning points and
each monotonic run drawn separately, with a warning naming the fold count.

The Standard Curve tab builds a model from known standards and predicts
unknowns, showing the fit and the plots curve_gen writes.

Distinct from the six-vial absorbance readout in Science Analysis, which
comes from the science embedded node over /science/sensor_data. Hidden by
default; capture additionally needs a caldata.txt and an attached camera.

Verified against Science_RD's real Rezasurin standards: R2 0.980833, slope
1.636312, matching their recorded yeast_model.json.
LayoutStore persists HMI layouts as JSON in src/rover_hmi_core/config/layouts
(resolved via ROVERFLAKE_ROOT with a build-time fallback), replacing
QSettings so layouts live in the repo. tiling_container hoists its tunables
and dedupes overlay/nav/persistence helpers. entrypoint and rover_env source
the ROS setup before building so a clean workspace has an environment.
ScienceModule gains drillmotorstatus, drillstepperdir and vacuumstatus, and
ScienceModule drives them with toggle and Lower/Hold/Raise controls that
reflect the published state. ScienceModulePipeline subscribes to both the
command and sensor topics so the pipeline view follows what was actually
commanded.

Adds runtime font zoom: Ctrl+= and Ctrl+- adjust, Ctrl+0 resets, cached in
/tmp so a scale survives an HMI restart without becoming permanent. A
FontScaleFilter reapplies the scale when a module rewrites a widget's
stylesheet at runtime. Baseline sizes are raised and XL/Title steps added
for touch-target controls.
Nothing in the repo publishes /science/sensor_data — the science embedded
node lives outside it — so every science module reads "--" forever and the
UI cannot be exercised without hardware.

ScienceSimModule publishes that topic at 10 Hz from the HMI itself, with
controls for drill lower/hold/raise, OSF flow, sample loaded and spectro
ready, so the pipeline and analysis views can be developed and demoed
against plausible data. Off by default; enable with the SIM toggle.
cv_bridge's compiled extension is built against the numpy 1.x ABI and
crashes under the container's numpy 2; raw rgb8/bgr8 conversion is a
few lines of frombuffer/tobytes, so drop the dependency entirely.
aaronrhim added a commit that referenced this pull request Aug 5, 2026
…e reviewed arm headers (PR #187 feedback)"

This reverts commit 020d3d3.
aaronrhim added a commit that referenced this pull request Aug 5, 2026
…re reviewed arm headers (PR #187 feedback)"

This reverts commit d5c5b0b.
aaronrhim added a commit that referenced this pull request Aug 5, 2026
@aaronrhim
aaronrhim force-pushed the feat/aaronrhim/hmi_deploy_layouts branch from 835989a to ac1e97a Compare August 5, 2026 06:25
…xporter

- homing reverted to main's convention: zero at switch, independent axes,
  park at idle_position; HOMED_WAITING/group retreat removed
- A5/A6 limit switches enabled (aux2 pin 0), wrist-aware set_axis_position
  stamps both differential motors; seed/overtravel/retreat respect
  homing_direction sign; M4 switch moved to pin 0
- homing UI: A5/A6 selectable, dialog matches switch-zero behavior
- dev_arm_moveit_config_v4: renamed self-references from stale v2 name
  (collided with the real v2 package)
…arm_driver_hardening branch

- homing state machine = main's (switch zero, A5 wrist special-case, A6 relative)
- raw BlockingCycle, no guardTransport/watchdog/bus-redetect (accepted main faults)
- calibration removed end to end: driver callbacks/stubs, HMI Calibrate buttons,
  /arm/calibration_request pub; HMI is RAM-only — Apply & Write Flash removed,
  write_flash dropped from MoteusConfigUpdate (flash/calibration are bench jobs)
- kept: telemetry contract (last-known + connected, can_device, motors_replying,
  masked limit switches), staggered conf-get reads (read-only), command clamps,
  A4 enabled, checkFaults dead code deleted
This reverts commit 00915e7.
Restores driver_process + the Start/Stop Driver button in motor telemetry
(as of 7f68038, before the heart-managed removal in 5f4d2ba). Test-only
branch — mainline uses heart_dev.yaml per PR #187 review.
export COLCON_DEFAULTS_FILE="${ROVERFLAKE_ROOT}/colcon.defaults.yaml"

# Rover operational DDS domain. Pre-exported values win (e.g. bench isolation).
export ROS_DOMAIN_ID="${ROS_DOMAIN_ID:-101}"

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why is this negative

Comment thread src/rover_gnss/gnss_data.md Outdated

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

idk if we need this

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Revert this file

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Revert this file

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Tabling for now. need to test on cbs

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is okay but should be specificied in the PR description. "Changed active joystick to yadda yadda"

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

in the future we should wrap all log msgs in a "#ifdef DEBUG_MSGS" to clean up terminal noise

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Remove this file

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

i try to avoid services just to reduce complexity but ehh isok

@aaronrhim
aaronrhim enabled auto-merge (squash) August 7, 2026 08:27
@aaronrhim
aaronrhim merged commit 85abbd9 into main Aug 7, 2026
3 checks passed
@aaronrhim
aaronrhim deleted the feat/aaronrhim/hmi_deploy_layouts branch August 7, 2026 08:30
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants