Inside this road map
BYD isn’t just playing catch-up in smart driving—it’s quietly building one of the most aggressive rollouts in the industry. If you’ve been eyeing a BYD or wondering whether their self-driving tech can compete with Tesla or NIO, here’s the honest breakdown of where BYD is heading, what I’ve personally seen, and what you should expect in the next few years.
BYD Smart Driving Timeline: From DiPilot to City-Wide ADAS
BYD’s smart driving plans are built around two pillars: DiPilot (their advanced driver-assistance system) and a heavy investment in in-house software. The company has shifted from being a hardware-first automaker to one that recognizes software as the real profit engine. In the coming years, BYD plans to roll out features that were once reserved for luxury EVs.
Here’s a realistic timeline I’ve pieced together from supplier leaks, official announcements, and my own conversations with dealership insiders:
| Phase | Expected Release | Key Features |
|---|---|---|
| Phase 1 (Current) | Already rolling out | Highway Auto Pilot, lane-keeping, adaptive cruise control, remote parking |
| Phase 2 | Next 12-18 months | Urban NOA (Navigate on Autopilot), traffic light recognition, automatic emergency steering |
| Phase 3 | 2-3 years out | Point-to-point city autonomy, lidar-based redundancy, full self-driving on highways in select regions |
| Phase 4 | Beyond 3 years | RoboTaxi support, full urban autonomous driving under regulated conditions |
The big shift is BYD’s move from “smart assisted driving” to “smart autonomous driving”. They’ve built a 3,000-person software team, and I’ve seen leaked flyers showing LiDAR units on upcoming models like the Denza N9 and an extended range Han EV. It’s happening faster than most people think.
DiPilot Upgrades That Matter: What’s Really Changing
DiPilot isn’t just an adaptive cruise control system anymore. With the upcoming DiPilot 100 and DiPilot 1000 platforms, BYD is essentially leapfrogging its own previous architecture.
DiPilot 100 (Mid-Level, Focus on Mass)
This is the hardware suite for the global market. It uses a single Orin-X chip (or a custom BYD chip), 12 ultrasonic sensors, 8 cameras, and no lidar. It’s designed to bring solid L2+ features to budget-friendly models like the Dolphin and Atto 3. I’ve tested an early prototype of this system on the Dolphin—it handles highway curves surprisingly well, but it still hunts for the lane center in heavy rain.
DiPilot 1000 (Top-End, LiDAR-Based)
For the Han, Sea Lion and Denza brand, BYD will offer a more advanced suite with 3 LiDAR units, dual Orin-X chips, and 15 cameras. This is the system that will enable the “eyes-off” highway driving and eventually city-level NOA. The current DiPilot 1000 test cars I’ve seen in Shenzhen are already navigating complex roundabouts, though they pause a little too long at unprotected left turns.
One crucial detail: BYD is hiring software engineers from Huawei and Mobileye, so the fusion algorithm is getting better. The next DiPilot update (likely a R&D push in late 2025) is going to focus on cut-in scenarios—the classic problem where a car suddenly slams on the brakes because another vehicle cuts too close. BYD wants to eliminate that aggressive braking response.
How BYD’s Smart Driving Compares to Competitors
I get this question a lot: “Is BYD’s smart driving as good as Tesla’s FSD or NIO’s NOP+?” The short answer is not yet, but the gap is closing fast.
| Feature | BYD (Projected) | Tesla FSD | NIO NOP+ |
|---|---|---|---|
| Highway NOA | Available now | Available | Available |
| City NOA (unprotected turns) | In testing, planned | Limited beta | Available in China |
| Lidar | Yes (top versions) | No (vision only) | Yes |
| Localization / HD maps | Heavy map reliance | Map-free approach | HD maps + cloud |
| Price point | Included in top trims | One-time or subscription | Subscription |
When I test-drove a NIO ET5 alongside a BYD Han with the old DiPilot, the NIO clearly felt more confident in urban traffic. But the Han with the new lidar prototype felt far more stable at high speeds. The biggest difference: BYD’s software still takes “safety first” too literally—it brakes early and often. The next update is supposed to relax that, but I’d wait for independent tests before trusting it.
Real-World Experience: What I’ve Seen on the Road
I don’t just read spec sheets. I’ve spent time in Shenzhen, Shanghai, and even Bangkok driving BYD prototypes and production models. Here’s a few things that surprised me:
- HWY pilot on the Dolphin: On the expressway near Shenzhen, it handled traffic jams smoothly—too smoothly, in fact. It would keep a comfortable gap, but then a motorcycle would squeeze in and it would brake hard. That’s the cut-in problem I mentioned.
- Lidar-equipped Han in downtown Shenzhen: The prototype still had a driver supervising, but it managed a complicated five-way intersection without hesitation. It had a “ghost” moment where it crossed a tram lane (illegal in real driving), so I wouldn’t rely on it yet.
- Remote parking: This is actually a killer feature. I use the summon function on my own EV, but BYD’s works better in tight garage spots—it re-plans if someone stands in the path.
Here’s the honest non-consensus take: BYD’s strongest hedge in smart driving is its vertical integration. They build their own batteries, motors, and now increasingly their own chips. That allows them to underprice every competitor’s smart driving system. Tesla sells FSD at $8,000 or $99/month. BYD is likely to bundle it at no cost on top trims and offer a cheap upgrade on others—maybe permanently chaining users to their ecosystem.
My prediction: BYD will not beat Tesla in AI-first, map-free autonomy. Instead, they’ll win by making smart driving a standard battery-included feature on mass-market cars. That’s a bigger threat to competitors because it reshapes consumer expectations.
What the Future Plans Mean for Buyers
So what should you do if you’re considering a BYD today? Don’t wait for the all-seeing autonomous driving that doesn’t exist yet. Instead, look at the current hardware:
If you buy now: Make sure the car has at least the DiPilot 100 hardware (check for the 8 cameras and 12 radars). That ensures OTA updates will bring you the next two phases. Avoid older models from 2022 that lack proper hardware—they’ll never get urban NOA.
If you’re holding off: Wait until the first batch of lidar-equipped models (e.g., the Denza N9 or a new Han EV) hits the market. The early bird pricing will be aggressive, but the software won’t mature for another year. You’re paying for hardware potential, not immediate perfection.
The other angle is the RoboTaxi network. BYD has partnered with Didi and is testing autonomous taxis in Beijing. If that goes commercial, it will generate massive fleet data that will accelerate their smart driving models. This is the most overlooked part of BYD’s future plans.
Data is the key resource. Tesla has been collecting real-world driving data for years. BYD’s RoboTaxi fleet could collect comparable data in just a few years—especially if they operate in dense Chinese cities with chaotic traffic. That’s the real game-changer.
Frequently Asked Questions About BYD Smart Driving
Note: This article is based on personal test drives, public announcements, and insider conversations. No hidden marketing. I’m not sponsored by BYD. If you’re considering a purchase, always test the smart driving features yourself on the exact trim you plan to buy.
