mirror of
https://github.com/sunnypilot/sunnypilot.git
synced 2026-10-01 08:13:44 +08:00
Ford: add opt-in bounded assistance for large turn entries
This commit is contained in:
Isaac Barham
@@ -0,0 +1,106 @@
|
||||
# Ford large-turn entry assist
|
||||
|
||||
The coordinated encoder can select the same C0/C1 packet even when its angle
|
||||
target increases: larger fields sometimes have the same predicted next state,
|
||||
and its tie break favors the smaller fields. This does not prove that the live
|
||||
PSCM responds identically. This opt-in trial adds a bounded command after that
|
||||
selection, making the turn-entry experiment observable at the CAN output.
|
||||
|
||||
## Enable and revert
|
||||
|
||||
In sunnylink Ford settings, keep **Selected-Action Path Tracking** and
|
||||
**Coordinated C0/C1 Steering** enabled, and **Model Geometry Reference** disabled.
|
||||
Enable **Large-Turn Entry Assist (Experimental)**, then make an offroad-to-onroad
|
||||
transition. The setting is default off and read once at startup. Disabling only
|
||||
the new setting and cycling offroad/onroad restores the previous coordinated
|
||||
controller. Disabling Selected-Action Path Tracking restores upstream Ford
|
||||
control. No hot switching or added steering-loop parameter I/O.
|
||||
|
||||
## Exactly what changes
|
||||
|
||||
The original desired steering angle, inverse, target-trend preview and centering
|
||||
trim remain. An additional command is weighted by three continuous ramps:
|
||||
|
||||
- Requested angle magnitude: zero through 30 degrees, full at 60 degrees.
|
||||
- Wheel lag into that turn: zero through 25 degrees, full at 50 degrees.
|
||||
- Growing filtered target rate: zero at steady/relaxing target, full at
|
||||
30 degrees/second.
|
||||
|
||||
Multiply the three weights. At full weight, add **0.60 m C0 and 0.04 rad C1**
|
||||
in the requested turn direction, then quantize and clip to the existing DBC
|
||||
bounds. These are trial tuning values, not recovered Ford constants or a
|
||||
guaranteed conversion to wheel angle. The 4× label refers to the earlier offline
|
||||
starting budget of 0.15 m / 0.01 rad, not a multiplication of all steering.
|
||||
|
||||
Only the extra term is suppressed on steeringPressed, fresh PSCM limit >=2,
|
||||
or a target already limited by the existing inverse acceleration allowance.
|
||||
Existing disengagement, fault and explicit override gates still govern the
|
||||
whole command. C2/C3 stay zero. No new integrator, retained maneuver state or
|
||||
encoder search is added. The observer continues consuming actual packed commands.
|
||||
|
||||
The extra goes to zero when the filtered target stops growing. **This does not
|
||||
clear the consequences of earlier commands:** the observer and encoder retain
|
||||
their state, so later requests can change. Normal path following and physical
|
||||
wheel release are not guaranteed unchanged. Delayed unwind is an accepted trial
|
||||
tradeoff, not claimed fixed.
|
||||
|
||||
## Why this strength
|
||||
|
||||
Nine recorded routes were compared: 183, 185, 17c, 166, 16a, 172, 177, 175 and
|
||||
17a. Each run used 556,776 input rows, actual adapter/packer calls, frozen model
|
||||
and vehicle measurements, and full available histories. Compare against the
|
||||
unchanged coordinated controller at `d203a109a`.
|
||||
|
||||
| Extra budget | Weak right 2: max total C0/C1 difference | Changed ordinary samples |
|
||||
|---|---|---|
|
||||
| 0.15 m / 0.01 rad | 0.26 m / 0.0185 rad | 2,923 / 236,385 (1.24%) |
|
||||
| 0.30 m / 0.02 rad | 0.41 m / 0.0285 rad | 2,071 / 236,385 (0.88%) |
|
||||
| **0.60 m / 0.04 rad** | **0.71 m / 0.0485 rad** | **1,493 / 236,385 (0.63%)** |
|
||||
|
||||
Total differences can exceed the immediate additive budget because earlier
|
||||
commands alter subsequent encoder decisions. The ordinary screen requires
|
||||
requests within ±30 degrees, engaged control, no touch/limitReached, and two
|
||||
seconds of continuously qualifying data on each side. It measures command
|
||||
differences, not lane error. Lower counts are not proof of better centering.
|
||||
|
||||
The 4× trial changes 76/334 and 104/326 samples in the two weak-right windows.
|
||||
Their peak absolute fields stay below 3.10 m C0 and 0.225 rad C1. Roundabout
|
||||
windows reach maximum differences of 0.63–0.66 m / 0.0425–0.0445 rad.
|
||||
|
||||
Nine of ten marked wobble windows match exactly. In route172's bookmark, 98/396
|
||||
samples change, up to 0.21 m C0 / 0.007 rad C1. Command total variation is slightly
|
||||
lower (C0 5.95→5.85 m; C1 0.3605→0.3545 rad); maximum individual steps are unchanged.
|
||||
This does not prove that the real wobble improves or remains equal.
|
||||
|
||||
Across all nine routes the trial touches a DBC field bound on 29 samples versus
|
||||
12 for baseline, including a short interval on175. It never exceeds those
|
||||
bounds. It changes the marked good unwind's channel split substantially:
|
||||
maximum C0/C1 differences of 2.16 m / 0.058 rad. That is not a measured wheel
|
||||
release delay. Larger requests reaching CAN do not establish more wheel motion.
|
||||
|
||||
## Verification and interpretation
|
||||
|
||||
Tests exercise the real Ford adapter and packer: startup/fallback, default-off
|
||||
parameter, both turn signs, no extra for small requests/errors, DBC clipping,
|
||||
target limits, overrides, fault gates, and clearing the added term while retaining
|
||||
the observer's real history. Full-route production verification compares both
|
||||
settings directly against the frozen 4× research output, including state and
|
||||
transmitted commands. UI source and generated schema are checked together.
|
||||
|
||||
The deployment gate additionally suppresses the term when the inverse target
|
||||
is acceleration-limited; no active extra in these nine research tapes coincided
|
||||
with that condition. This does not introduce a new acceleration allowance.
|
||||
|
||||
`Ford joint path tracking` diagnostics identify the enabled experiment as
|
||||
`ford-joint-turn-entry-v1`, with `turn_entry_enabled`, `turn_entry_weight`,
|
||||
`turn_entry_c0`, `turn_entry_c1`, `base_wire_command` and actual `wire_sent`.
|
||||
The existing `predicted_curvature` field describes the encoder prediction before
|
||||
the added term; `filtered_curvature` is the observer advanced from actual sends.
|
||||
Logging cadence is unchanged.
|
||||
|
||||
These are offline numerical checks, not a closed-loop steering validation.
|
||||
166/16a retain raw-yaw sensitivity; 183 has incomplete earlier history; 185 has
|
||||
log gaps. The recovered older firmware model is not proven identical to the
|
||||
live truck. The physical question for this trial is whether the stronger entry
|
||||
requests move the wheel farther/faster without unacceptable normal-driving
|
||||
changes. Neither turn completion nor preserved driving feel is established.
|
||||
@@ -242,6 +242,7 @@ inline static std::unordered_map<std::string, ParamKeyAttributes> keys = {
|
||||
{"FordC0TimeBased", {PERSISTENT | BACKUP, BOOL, "0"}},
|
||||
{"FordGeometryReference", {PERSISTENT | BACKUP, BOOL, "0"}},
|
||||
{"FordPscmJointControl", {PERSISTENT | BACKUP, BOOL, "0"}},
|
||||
{"FordPscmTurnEntryAssist", {PERSISTENT | BACKUP, BOOL, "0"}},
|
||||
{"HyundaiLongitudinalTuning", {PERSISTENT | BACKUP, INT, "0"}},
|
||||
{"SubaruStopAndGo", {PERSISTENT | BACKUP, BOOL, "0"}},
|
||||
{"SubaruStopAndGoManualParkingBrake", {PERSISTENT | BACKUP, BOOL, "0"}},
|
||||
|
||||
@@ -9,12 +9,23 @@ import math
|
||||
from opendbc.car.ford.values import FordFlags
|
||||
|
||||
# Small steering-wheel trim around the nominal inverse, in degrees. These are
|
||||
# trial tuning values, not recovered PSCM constants. No extra proportional loop.
|
||||
# trial tuning values, not recovered PSCM constants. Entry assist is separately gated.
|
||||
ANGLE_TRIM_MAX = 2.0
|
||||
ANGLE_TRIM_KI = 0.2 # 1/s; error clipping limits adaptation to 0.4 deg/s.
|
||||
ANGLE_TRIM_ERROR_MAX = 5.0
|
||||
ANGLE_TRIM_RATE_MAX = 5.0 # deg/s: do not learn a bias from fast wheel motion.
|
||||
TARGET_PREVIEW = 0.1 # seconds; causal action-trend forecast, then hold.
|
||||
TURN_ENTRY_C0_MAX = 0.60 # metres; experimental post-encoder correction budget.
|
||||
TURN_ENTRY_C1_MAX = 0.04 # radians; not a recovered Ford calibration value.
|
||||
|
||||
|
||||
def turn_entry_weight(target, angle, requested_rate):
|
||||
"""Fade in only for a large, growing turn request with substantial wheel lag."""
|
||||
direction = math.copysign(1.0, target)
|
||||
size = max(0.0, min(1.0, (abs(target) - 30.0) / 30.0))
|
||||
lag = max(0.0, min(1.0, (direction * (target - angle) - 25.0) / 25.0))
|
||||
growth = max(0.0, min(1.0, direction * requested_rate / 30.0))
|
||||
return size * lag * growth
|
||||
|
||||
|
||||
def joint_control_enabled(CP, params):
|
||||
@@ -30,17 +41,18 @@ def joint_control_enabled(CP, params):
|
||||
|
||||
def select_joint_control(CP, params):
|
||||
# No calibration or native-library load on the normal/default path.
|
||||
return FordJointControl(CP) if joint_control_enabled(CP, params) else None
|
||||
return FordJointControl(CP, turn_entry_assist=params.get_bool('FordPscmTurnEntryAssist')) if joint_control_enabled(CP, params) else None
|
||||
|
||||
|
||||
class FordJointControl:
|
||||
def __init__(self, CP):
|
||||
def __init__(self, CP, *, turn_entry_assist=False):
|
||||
from opendbc.can import CANParser
|
||||
from opendbc.car.ford.fordcan import CanBus
|
||||
from openpilot.selfdrive.controls.lib.ford_joint.model import MainRequest
|
||||
from openpilot.selfdrive.controls.lib.ford_joint.angle import AngleModel
|
||||
from openpilot.selfdrive.controls.lib.ford_joint.encoder import PairedRelease
|
||||
|
||||
self.turn_entry_assist = turn_entry_assist
|
||||
self.wheelbase, self.ratio = CP.wheelbase, CP.steerRatio
|
||||
if not all(math.isfinite(v) and v > 0 for v in (self.wheelbase, self.ratio)):
|
||||
raise ValueError('Joint control requires finite positive vehicle geometry')
|
||||
@@ -80,7 +92,7 @@ class FordJointControl:
|
||||
self.last_time = now
|
||||
|
||||
def prepare(self, CC, CC_SP, CS, now, *, fresh=True, pscm_status=None):
|
||||
from openpilot.selfdrive.controls.lib.ford_joint.inverse import C0_BOUND, C1_BOUND, invert_angle
|
||||
from openpilot.selfdrive.controls.lib.ford_joint.inverse import C0_BOUND, C1_BOUND, invert_angle, quantize
|
||||
|
||||
dt = now - self.last_time if self.last_time is not None else 0.0
|
||||
self.advance(now)
|
||||
@@ -150,7 +162,21 @@ class FordJointControl:
|
||||
curvature_rate=self.requested_rate / inverse['slope_per_curvature'], preview=TARGET_PREVIEW,
|
||||
curvature_bound=inverse['allowance'] / (speed / 3.6)**2,
|
||||
)
|
||||
base_command = command
|
||||
entry_weight = 0.0
|
||||
if self.turn_entry_assist and not CS.steeringPressed and not inverse['accel_limited'] and not (status_fresh and pscm_status.limit >= 2):
|
||||
entry_weight = turn_entry_weight(target, angle, self.requested_rate)
|
||||
if entry_weight:
|
||||
extra = math.copysign(entry_weight, target)
|
||||
command = quantize((command[0] + extra * TURN_ENTRY_C0_MAX, command[1] + extra * TURN_ENTRY_C1_MAX))
|
||||
# Apply after selection so nominal slew equivalence cannot discard the
|
||||
# added request. record_sent/advance still observe the actual packets;
|
||||
# their retained state can affect later commands after this term clears.
|
||||
details = {
|
||||
'base_wire_command': tuple(map(float, base_command)),
|
||||
'turn_entry_weight': entry_weight,
|
||||
'turn_entry_c0': float(command[0] - base_command[0]),
|
||||
'turn_entry_c1': float(command[1] - base_command[1]),
|
||||
'requested_angle': target,
|
||||
'trimmed_angle': trimmed_target,
|
||||
'angle_trim': self.angle_trim,
|
||||
@@ -161,7 +187,7 @@ class FordJointControl:
|
||||
'target_preview_limited': info['preview_limited'],
|
||||
'reachable_angle': inverse['reachable_target'],
|
||||
'target_curvature': inverse['curvature'],
|
||||
'predicted_curvature': float(info['first_state'][3]),
|
||||
'predicted_curvature': float(info['first_state'][3]), # Encoder prediction before optional entry assist.
|
||||
'accel_limited': inverse['accel_limited'],
|
||||
}
|
||||
# Learn only small residual errors while the wheel is settling. Large
|
||||
@@ -189,7 +215,11 @@ class FordJointControl:
|
||||
cc = CC.as_reader().as_builder() if hasattr(CC, 'as_reader') else CC.as_builder()
|
||||
cc.latActive = active
|
||||
self.diagnostics = {
|
||||
'hypothesis': 'ford-joint-v24',
|
||||
'hypothesis': 'ford-joint-turn-entry-v1' if self.turn_entry_assist else 'ford-joint-v24',
|
||||
'turn_entry_enabled': self.turn_entry_assist,
|
||||
'turn_entry_weight': 0.0,
|
||||
'turn_entry_c0': 0.0,
|
||||
'turn_entry_c1': 0.0,
|
||||
'status': self.fault or ('active' if active else 'inactive'),
|
||||
'driver_override': override,
|
||||
'driver_pressed': bool(CS.steeringPressed),
|
||||
|
||||
@@ -0,0 +1,110 @@
|
||||
"""Opt-in entry assist at the real Ford adapter and CAN packing boundary."""
|
||||
from types import SimpleNamespace
|
||||
|
||||
import numpy as np
|
||||
import pytest
|
||||
|
||||
from openpilot.common.params import Params, ParamKeyFlag
|
||||
from openpilot.selfdrive.car.ford_joint_control import FordJointControl, select_joint_control
|
||||
from openpilot.selfdrive.car.tests.test_ford_joint_control import Pipeline, cp
|
||||
from openpilot.selfdrive.controls.lib.ford_joint.encoder import state
|
||||
|
||||
|
||||
def assisted():
|
||||
p = Pipeline()
|
||||
p.joint = FordJointControl(cp(), turn_entry_assist=True)
|
||||
return p
|
||||
|
||||
|
||||
def test_turn_entry_is_default_off_and_requires_the_existing_joint_controller(tmp_path):
|
||||
params = Params(str(tmp_path))
|
||||
assert params.get_default_value('FordPscmTurnEntryAssist') is False
|
||||
assert not params.get_bool('FordPscmTurnEntryAssist')
|
||||
assert b'FordPscmTurnEntryAssist' in params.all_keys(ParamKeyFlag.PERSISTENT | ParamKeyFlag.BACKUP)
|
||||
params.put_bool('FordPscmTurnEntryAssist', True, block=True)
|
||||
assert select_joint_control(cp(), params) is None
|
||||
params.put_bool('FordModelActionController', True, block=True)
|
||||
assert select_joint_control(cp(), params) is None
|
||||
params.put_bool('FordPscmJointControl', True, block=True)
|
||||
assert select_joint_control(cp(), params).turn_entry_assist
|
||||
params.put_bool('FordPscmTurnEntryAssist', False, block=True)
|
||||
assert not select_joint_control(cp(), params).turn_entry_assist
|
||||
params.put_bool('FordPscmTurnEntryAssist', True, block=True)
|
||||
for key in ('FordGeometryReference', 'JoystickDebugMode'):
|
||||
params.put_bool(key, True, block=True)
|
||||
assert select_joint_control(cp(), params) is None
|
||||
params.put_bool(key, False, block=True)
|
||||
|
||||
|
||||
@pytest.mark.parametrize('sign', [-1., 1.])
|
||||
def test_large_growing_undertracked_request_adds_to_real_wire_commands(sign):
|
||||
base, trial = Pipeline(), assisted()
|
||||
for p in (base, trial):
|
||||
p.tick(1., 0.)
|
||||
p.tick(1.01, sign * 120.)
|
||||
np.testing.assert_allclose(np.array(trial.joint.sent[:2]) - base.joint.sent[:2], sign * np.array([.60, .04]), atol=1e-12)
|
||||
assert trial.joint.diagnostics['requested_angle'] == sign * 120.
|
||||
assert trial.joint.diagnostics['trimmed_angle'] == base.joint.diagnostics['trimmed_angle']
|
||||
assert trial.joint.diagnostics['requested_rate'] == base.joint.diagnostics['requested_rate']
|
||||
assert trial.joint.diagnostics['turn_entry_weight'] == 1.
|
||||
assert trial.joint.sent[2] and base.joint.sent[2]
|
||||
|
||||
|
||||
@pytest.mark.parametrize('sign', [-1., 1.])
|
||||
def test_extra_clips_at_dbc_bounds_through_real_packer(sign, monkeypatch):
|
||||
p = assisted()
|
||||
p.tick(1., 0.)
|
||||
choose = p.joint.encoder.choose
|
||||
|
||||
def near_bounds(*args, **kwargs):
|
||||
_, info = choose(*args, **kwargs)
|
||||
return (sign * 5.10, sign * .49), info
|
||||
|
||||
monkeypatch.setattr(p.joint.encoder, 'choose', near_bounds)
|
||||
p.tick(1.01, sign * 120.)
|
||||
assert p.joint.sent[:2] == pytest.approx((sign * 5.11, sign * .5))
|
||||
|
||||
|
||||
def test_small_requests_and_small_errors_keep_full_history_identical():
|
||||
base, trial = Pipeline(), assisted()
|
||||
for i in range(1200):
|
||||
target = 25 * np.sin(i * .02) if i < 600 else 100.
|
||||
wheel = 0. if i < 600 else 90.
|
||||
for p in (base, trial):
|
||||
p.cs.steeringAngleDeg = wheel
|
||||
p.tick(1. + i * .01, float(target))
|
||||
assert trial.joint.sent == base.joint.sent
|
||||
np.testing.assert_array_equal(state(trial.joint.request), state(base.joint.request))
|
||||
assert trial.joint.angle_trim == base.joint.angle_trim
|
||||
|
||||
|
||||
@pytest.mark.parametrize('sign', [-1., 1.])
|
||||
def test_new_term_clears_on_relaxation_without_resetting_observed_state(sign):
|
||||
p = assisted()
|
||||
for i in range(100):
|
||||
p.cs.steeringAngleDeg = sign * 20.
|
||||
p.tick(1. + i * .01, sign * (100. + i))
|
||||
assert p.joint.diagnostics['turn_entry_c0'] * sign > 0
|
||||
p.tick(2., sign * 50.)
|
||||
assert p.joint.diagnostics['turn_entry_c0'] == p.joint.diagnostics['turn_entry_c1'] == 0.
|
||||
assert abs(p.joint.request.c0) + abs(p.joint.request.c1) > 0
|
||||
assert p.joint.sent[2]
|
||||
|
||||
|
||||
@pytest.mark.parametrize('gate', ['touch', 'limit', 'denied', 'fault', 'stale', 'inactive', 'accel'])
|
||||
def test_extra_respects_intervention_health_and_target_limit_gates(gate):
|
||||
p = assisted()
|
||||
p.cs.vEgo = 30. if gate == 'accel' else 5.36
|
||||
p.tick(1., 0.)
|
||||
p.cs.steeringPressed = gate == 'touch'
|
||||
p.cs.steerFaultTemporary = gate == 'fault'
|
||||
status = SimpleNamespace(valid=True, canMonoTime=1_010_000_000, limit=2 if gate == 'limit' else 0, denied=gate == 'denied')
|
||||
p.tick(1.01, 120., active=gate != 'inactive', fresh=gate != 'stale', pscm_status=status)
|
||||
d = p.joint.diagnostics
|
||||
assert d['turn_entry_c0'] == d['turn_entry_c1'] == 0.
|
||||
if gate in ('touch', 'limit', 'accel'):
|
||||
assert p.joint.sent[2] # Only the added term is suppressed.
|
||||
else:
|
||||
assert p.joint.sent == (0., 0., False)
|
||||
if gate == 'accel':
|
||||
assert d['accel_limited']
|
||||
@@ -2231,6 +2231,34 @@
|
||||
"equals": false
|
||||
}
|
||||
]
|
||||
},
|
||||
{
|
||||
"key": "FordPscmTurnEntryAssist",
|
||||
"widget": "toggle",
|
||||
"needs_onroad_cycle": true,
|
||||
"title": "Large-Turn Entry Assist (Experimental)",
|
||||
"description": "Add bounded steering correction when a large turn request is growing and the wheel is behind.",
|
||||
"details": "Requires Coordinated C0/C1 Steering and Selected-Action Path Tracking on, with Model Geometry Reference off. Default off. Leaves the normal steering target and centering logic in place. This trial may increase turn authority but can affect later centering and hold a turn longer; physical improvement has not been established. Turning it off returns to the existing coordinated controller after an offroad-to-onroad cycle. Turning Selected-Action Path Tracking off restores upstream Ford control.",
|
||||
"enablement": [
|
||||
{
|
||||
"type": "offroad_only"
|
||||
},
|
||||
{
|
||||
"type": "param",
|
||||
"key": "FordModelActionController",
|
||||
"equals": true
|
||||
},
|
||||
{
|
||||
"type": "param",
|
||||
"key": "FordPscmJointControl",
|
||||
"equals": true
|
||||
},
|
||||
{
|
||||
"type": "param",
|
||||
"key": "FordGeometryReference",
|
||||
"equals": false
|
||||
}
|
||||
]
|
||||
}
|
||||
]
|
||||
},
|
||||
|
||||
@@ -43,6 +43,23 @@ sections:
|
||||
- type: param
|
||||
key: FordGeometryReference
|
||||
equals: false
|
||||
- key: FordPscmTurnEntryAssist
|
||||
widget: toggle
|
||||
needs_onroad_cycle: true
|
||||
title: Large-Turn Entry Assist (Experimental)
|
||||
description: Add bounded steering correction when a large turn request is growing and the wheel is behind.
|
||||
details: Requires Coordinated C0/C1 Steering and Selected-Action Path Tracking on, with Model Geometry Reference off. Default off. Leaves the normal steering target and centering logic in place. This trial may increase turn authority but can affect later centering and hold a turn longer; physical improvement has not been established. Turning it off returns to the existing coordinated controller after an offroad-to-onroad cycle. Turning Selected-Action Path Tracking off restores upstream Ford control.
|
||||
enablement:
|
||||
- $ref: '#/macros/offroad'
|
||||
- type: param
|
||||
key: FordModelActionController
|
||||
equals: true
|
||||
- type: param
|
||||
key: FordPscmJointControl
|
||||
equals: true
|
||||
- type: param
|
||||
key: FordGeometryReference
|
||||
equals: false
|
||||
- id: hyundai
|
||||
title: Hyundai / Kia / Genesis Settings
|
||||
description: ''
|
||||
|
||||
@@ -317,6 +317,17 @@ class TestKnownVehicleSettings(OpenpilotTestCase):
|
||||
with tempfile.TemporaryDirectory() as path:
|
||||
assert Params(path).get_default_value("FordPscmJointControl") is False
|
||||
|
||||
def test_ford_turn_entry_is_separate_default_off_cycle_only_trial(self, schema):
|
||||
items = _brand_items(schema["vehicle_settings"].get("ford"))
|
||||
item = next(item for item in items if item["key"] == "FordPscmTurnEntryAssist")
|
||||
assert item["widget"] == "toggle" and item["needs_onroad_cycle"] is True
|
||||
assert item["enablement"] == [{"type": "offroad_only"},
|
||||
{"type": "param", "key": "FordModelActionController", "equals": True},
|
||||
{"type": "param", "key": "FordPscmJointControl", "equals": True},
|
||||
{"type": "param", "key": "FordGeometryReference", "equals": False}]
|
||||
with tempfile.TemporaryDirectory() as path:
|
||||
assert Params(path).get_default_value("FordPscmTurnEntryAssist") is False
|
||||
|
||||
def test_hyundai_has_longitudinal_tuning(self, schema):
|
||||
keys = {i["key"] for i in _brand_items(schema["vehicle_settings"].get("hyundai"))}
|
||||
assert "HyundaiLongitudinalTuning" in keys
|
||||
|
||||
Reference in New Issue
Block a user