On October 14, 2026, during the Paris Motor Show, He Xiaopeng, Chairman and CEO of XPENG Group, attended the Paris Automotive Summit, released the smart vehicle solution XPENG AI ENGINEERING, and announced progress in the physical AI ecosystem. The solution consists of a vehicle platform, modular technology, and a toolchain; the company claims it saves 41 minutes in per-vehicle inspection time and reduces annual operating costs of production-line electrical inspection by 95%. Timelines for three product lines were also disclosed: humanoid robots are planned for mass production by the end of 2026 and delivery in the first half of 2027; Robotaxi has been given the Chinese and English names “小鹏悠游” and “XPENGYOYO” and is in talks with European mobility platforms; flying cars have cumulative orders exceeding 7,000 units, and NGP is planned to be gradually delivered to global customers starting in 2027. XPENG’s role in Europe is extending from a complete-vehicle exporter to a technology supplier, while the pace of implementation and local compliance remain to be verified.
On October 14, 2026, during the Paris Motor Show, XPENG Group Chairman and CEO He Xiaopeng attended the Paris Automotive Summit as the only invited CEO of a Chinese automaker, delivered a speech, and unveiled the smart vehicle solution “XPENG AI ENGINEERING,” while also announcing progress and a cooperation framework for the physical AI ecosystem. Unlike previous Chinese automakers at European auto shows, which focused their launches on new models, XPENG this time put “technology openness” and “ecosystem cooperation” first.
From product exports to technology openness
According to disclosures, XPENG AI ENGINEERING consists of three parts: the vehicle platform, modular technology, and toolchain. The vehicle platform covers platform architecture and electronic/electrical architecture, with bandwidth spanning compact to large vehicles and the ability to improve parts reuse; the electronic/electrical architecture adopts 10-gigabit Ethernet, increases in-vehicle transmission speed by 33 times, and supports steer-by-wire chassis and L3-oriented safety redundancy.
Modular technology involves intelligent driving, AI BOX, chips, batteries, and electric drive: Turing AI intelligent driving is equipped with the second-generation VLA large model; AI BOX can run 30-billion-parameter MoE large models; the 8C ultra-fast-charging battery can increase charge from 10% to 80% in 8 minutes and supports 15-minute full-vehicle OTA. The toolchain covers calibration, EOL, AI diagnostics, and other steps. Official figures show that cumulative inspection time per vehicle is reduced by 41 minutes and annual operating costs for production-line electrical inspection are lowered by 95%.
Opening this solution to global partners means XPENG’s role in Europe is extending from complete-vehicle exporter to technology supplier. For overseas automakers, the appeal lies in shortening development cycles and lowering production-line costs; whether cooperation can be reached depends on localization adaptation, data compliance, and long-term service capabilities, for which no public details are available.
Timelines for three product lines
The physical AI ecosystem covers humanoid robots, Robotaxi, flying cars, Turing AI chips, and physical-world foundation models. The timelines for the three product lines were presented side by side for the first time in this speech:
- Humanoid robots: planned for mass production by the end of 2026 and delivery in the first half of 2027.
- Robotaxi: in talks with multiple European mobility platforms, and adopting the Chinese name “小鹏悠游” and the English name “XPENGYOYO.”
- Flying cars: cumulative orders exceed 7,000 units.
In intelligent driving, XPENG NGP has completed its first large-scale public test in Europe and plans to gradually deliver to global customers starting in 2027. From this, intelligent driving, Robotaxi, and robots share the same AI capability foundation, with Robotaxi being the form earliest integrated with overseas operational scenarios.
Europe deployment path and unresolved issues
From the timeline, XPENG has set the end of 2026 to the first half of 2027 as a concentrated delivery window: robot mass production and delivery are only about half a year apart, and global delivery of NGP also starts in 2027. This pace poses simultaneous challenges to the supply chain, production capacity, and local service networks.
Second, the deployment logic for the three categories of products is not the same. Robotaxi relies on cooperation with European mobility platforms, involving operating entities, liability allocation, and local regulatory permits; flying cars already have cumulative orders of more than 7,000 units, but converting orders into deliveries still requires airworthiness certification and supporting operating scenarios; for humanoid robots, mass production capability is the main variable, and commercialization scenarios were not elaborated in this disclosure.
In addition, the advancement of open technology cooperation in the European market depends on the extent to which local partners accept the platform and toolchain. This disclosure did not cover the specific European countries where the above-mentioned products will be launched, production arrangements, or the scale of investment, nor did it provide progress milestones for the Robotaxi negotiations.
For Chinese automakers, what XPENG presented in Paris this time is an overseas expansion framework centered on AI capabilities and covering both whole-vehicle technology export and multi-form physical AI products. The value of the framework will ultimately still depend on whether delivery can be fulfilled: whether the humanoid robot can be delivered on schedule in the first half of 2027, whether Robotaxi can finalize its first operating partnership in Europe, and whether the brand name "XPENG Youyou" can truly be implemented as a service that passengers can perceive—all of these still require subsequent data for verification.
