Changsha's AI Robot Cooks Pepper-Pork Stir-Fry in 2:17, Recreating Wok Hei

In the Xinchushi showroom at Xiangjiang Sci-Tech Innovation Base, Yuelu District, Changsha, three cooking robots are running at full speed. They produce a plate of pepper-pork stir-fry in 2 minutes 17 seconds and a stir-fried kidney dish in just 1 minute 30 seconds. As soon as the steaming Hunan dishes are placed on the counter, restaurant managers gather around to taste them.

Operators simply load pre-prepared ingredients into the hopper; the robots handle the rest automatically—feeding, stir-frying, adjusting heat, and plating—all following built-in digital recipes. Each recipe can be flexibly adjusted according to regional tastes.

"We decompose a chef's craft into data and turn family recipes into replicable standard procedures, achieving 'one taste across a thousand stores,'" says Liao Hui, founder of Xinchushi. The team developed AI temperature-curve approximation algorithms, AI timing recognition, and a molecular-level seasoning system to precisely replicate a chef's heat control, wok hei, and authentic flavor.

In the past, chef phrases like "70% oil temperature" or "a pinch of salt" relied on feel and couldn't be quantified. Now every ingredient addition, seasoning amount, and heat change is accurately recorded in digital recipes.

Chinese restaurant chains have long faced a hurdle in cross-regional expansion: flavor inconsistency. Industry data shows that over 60% of Chinese chain brands lose customers to unstable taste when opening new locations. Now, this challenge is being cracked by Changsha local AI enterprises.