- Overview: Why This Update Matters
- What’s New in the BYD Smart Driving Update
- Highway Assist Test: 500 Miles in Real Traffic
- Parking Evolution: From Hesitant to Confident
- How Did the Driver Monitoring Nags Change?
- How Does This Update Compare to Competitors?
- Known Bugs and Workarounds (I Found 3)
- Frequently Asked Questions
Overview: Why This Update Matters
I’ve been driving a BYD Han EV with the DiPilot system for eight months. When the OTA notification for a “smart driving update” popped up, I was skeptical. Previous updates were minor—smoother lane keeping, slightly better curve handling. But this one? The release notes mentioned “highway assist enhancement” and “automatic parking optimization.”
I decided to put it through a real test: a 500-mile round trip from Shanghai to Nanjing, split across highway, suburban roads, and a few tricky parking lots. I wanted to see if BYD had finally caught up with the more mature systems from NIO or XPeng—or if this was just another incremental patch.
What’s New in the BYD Smart Driving Update
According to the official changelog (v2.3.0, received mid-month), the update focuses on three areas:
- Highway Assist (HA): Improved lane centering in curves, better response to cut-ins, and reduced false braking.
- Traffic Jam Pilot (TJP): Faster acceleration after stop-and-go, smoother stop timing.
- Automatic Parking: Support for parallel parking in tighter spots, enhanced detection of low obstacles like curbs.
But changelogs never tell the full story. Let me walk you through what I actually experienced.
Highway Assist Test: 500 Miles in Real Traffic
I started early morning to catch moderate traffic. On the G42, with light traffic, the car held its lane beautifully—much less micro-correction than before. In previous versions, the steering wheel would jitter every few seconds; now it’s almost imperceptible. I could actually relax my grip.
The real challenge came near Suzhou: construction zones with narrow lanes and erratic merging. The system identified cut-ins earlier and didn’t brake abruptly. Instead, it gradually reduced speed, which felt more human. I noticed it still hesitates about half a second before reacting—enough to make me pay attention, but not alarming.
One specific improvement: on a long, gentle left curve (radius ~500 meters), the previous software would sometimes drift to the right lane line before correcting. This update kept the car dead center. I checked with the lane departure indicator—no red flashes.
Personal observation: The system now handles rain-slicked roads better. I hit a brief downpour near Kunshan, and the HA didn’t disengage. It reduced speed by about 10 km/h automatically, something I’ve never seen in any NIO or Tesla I’ve driven.
But it’s not perfect. On sharp curves (advisory speed below 60 km/h), the car still brakes too late for my taste. It enters the curve at set speed, then brakes midway. I’d rather it slow down before the turn. BYD’s algorithm seems optimized for comfort, not confidence.
Parking Evolution: From Hesitant to Confident
Automatic parking was my biggest pain point before. The old system would abort if a leaf blew near the sensor. This update? I tested it in three scenarios:
| Scenario | Old Behavior | New Behavior |
|---|---|---|
| Parallel spot (4.8m gap) | Aborted after first reverse | Completed in 3 moves (took 45s) |
| Perpendicular spot (with car on one side) | Parked but off-center ~20cm | Centered within 5cm |
| Tight spot (curb on right, wall on left) | Hit the curb (scratched rim) | Detected curb and left extra space (10cm) |
I even tried a spot with a low concrete barrier (15cm tall). The new update recognized it and stopped 30cm away—old system would have bumped. The parking speed is still slow, but I prefer cautious over fast and hit.
How Did the Driver Monitoring Nags Change?
BYD uses an infrared camera above the steering column. In the older software, if I glanced at the nav screen for more than 3 seconds, it would chime annoyingly. Now the threshold seems relaxed: about 5 seconds. That’s good, because the navigation on the central display is actually useful.
However, the hands-on detection remains strict. If I rest my hand on the wheel at 7 o’clock instead of 9 and 3, it nags after 15 seconds. I get why—safety—but on a straight highway, I’d love a lighter touch. No change there.
How Does This Update Compare to Competitors?
I’ve driven the NIO ET5 with NOP+ and the XPeng P7 with NGP. Here’s my honest take:
- Lane keeping: BYD now ties with XPeng—smooth but still conservative. NIO is slightly more aggressive.
- Curve handling: XPeng leads (no braking on tight curves). BYD improved but not enough.
- Cut-in response: BYD now comparable to NIO—gradual deceleration vs. sudden braking of earlier versions.
- Parking: BYD beats NIO in parallel parking accuracy but still slower than XPeng.
- Driver monitoring: XPeng’s is most lenient; BYD is in the middle.
Where BYD falls short is in navigation integration. The system doesn’t read speed limit signs consistently; it relies on map data that’s sometimes outdated. On a road with a recent speed limit change from 80 to 60 km/h, the car didn’t slow down until I manually adjusted. XPeng and NIO both update from cameras and maps.
Known Bugs and Workarounds (I Found 3)
No update is perfect. Here are three issues I encountered and how to handle them:
- Auto lane change cancelled randomly: On a two-lane highway, the car started a lane change to the left, then aborted halfway because a truck was approaching fast—but the sensor saw it late. The car swerved back abruptly. Workaround: Only use auto lane change when you have clear visibility; manually cancel if you sense danger.
- Navigation-assisted cruise control doesn’t slow for toll booths: The car maintains set speed until the last moment—then brakes hard. I almost rear-ended a queue. Workaround: Disengage cruise well before toll plaza.
- Automatic parking fails in fading light: At dusk, the system couldn’t detect a white line on light gray asphalt. It just stopped and asked me to park manually. Workaround: Ensure good lighting or park manually.
Reader Comments