SMARTWELL

巡检 采摘 打叶 授粉
温室与作物为实景 · 机器人为研发阶段行业示意
SYSTEM ONLINE · R&D
IK SOLVE 60 Hz
FORCE ±0.1 N · GRIP QUICK-SWAP
VISION MULTISPECTRAL · FLEET SWARM

AI model as the brain, robots as the arms.

走进温室——四大作业,同一个大脑

机器人四大作业的真实温室场景,叠加模型实时读取的数据。

06:30 · SCAN NDVI 0.87 · 22.6°C

清晨巡检

沿垄多光谱巡检,人还没上班,异常已经报到大脑。

14:00 · HARVEST ±0.1N · damage <3%

轻柔采摘

高架槽上力控采摘熟果——不碰伤。

18:30 · PRUNE leaf count −12%/day

打叶疏枝

末端几分钟换装——熄灯前完成打叶与打顶。

10:00 · POLLINATE fruit set +10%

振动授粉

振动棒按计划巡花——不依赖蜂群。

照片说明:温室与作物为实景;机器人为行业参考示意,自研机器人在研。

A day in the life of one robot

06:30

Morning scout

Multispectral pass down every row; anomalies flagged to the brain before staff arrive.

10:00

Precision fertigation

Variable-rate nutrients per zone: fertigation error <5%, irrigation ±3%.

14:00

Gentle harvest

±0.1N force control picks ripe berries without bruising; damage <3%.

18:30

De-leaf &amp; pollinate

End effectors swap in minutes — de-leafing, topping, pollination before lights dim.

Five robots, one brain

Inspection Robot

Scouts crops 24/7 with vision + multispectral: 95%+ anomaly detection, NPK probes at <5% error, growth alerts to the brain.

全作物 / All crops

Precision-Operation Robot

Variable-rate fertigation and spraying ends, drip/flood heads, dynamic recipes — feeding each zone exactly what it needs.

番茄 · 叶菜 / Tomatoes · Greens

Harvesting Robot

Berry- and apple-adaptive picking ends, de-leafing, topping, pollination — ±0.1N force, <3% damage.

草莓 · 番茄 / Berries · Tomatoes Flagship

Logistics Robot

Carries fruit bins and tanks, crosses rows and ridges adaptively — the supply line of the fleet.

果园 · 温室 / Orchards · Greenhouses

Climate-Regulation Robot

Mobile sensing plus spot climate actions inside CEA — closing the loop between decisions and actuators.

温室 / Greenhouses

Shared Agri-Brain

All five robots run on the same dual-engine crop models.

Six core sub-modules

Environment-Sensing End

Light / air-humidity sensing heads · soil NPK & pH rapid probes · fruit/leaf vision and multispectral scouting sensors · growth-anomaly early warning.

Precision Fertigation End

Variable-rate fertilizer/spray ends · precision drip/flood heads · fertigation pipeline control · on-demand dynamic recipes.

Crop-Management End

Picking ends (berry/apple-adaptive) · de-leafing and topping ends · pollination ends — full coverage of field operations.

Robotic Arm Sub-module

Multi-DOF intelligent arm · high-precision adaptive joints · fast-response servos · quick-swap interfaces for multiple end types.

Multi-Robot Coordination

Monitoring / fertigation / care multi-task scheduling · multi-arm collision avoidance and production planning · fleet throughput scaling.

Stable Field Platform

High-stability damped chassis · lift/traverse adjustment · tank and fruit-bin integration · cross-row adaptive traversal.

95%+ Anomaly Detection Accuracy
±0.1N Force Control Precision
<3% Harvest Damage Rate
6× Daily Efficiency vs Manual
//Regions / structures / orchards validated

From lab validation to commercial fleet

2026Q3—2027Q2 · Phase 1 · Validation

Core module R&D in lab & pilot farms 进行中

Adaptive arm joints, force control (±0.1N) and quick-swap end effectors validated across 3 production regions, 8 growing structures and 5 demo orchards.

2027Q3—Q4 · Phase 2 · Prototype

First integrated field platform 规划中

Mobile chassis + multi-DOF arm + vision stack integrated; supervised picking, de-leafing and pollination trials in CEA greenhouses.

2028Q1-Q3 · Phase 3 · Pilot fleet

Multi-robot scheduling pilot 规划中

Fleet-level task scheduling and arm-arm collision avoidance; pilot deployments with anchor customers (Nanjing cooperation in planning).

2028 Q4 · Phase 4 · Scale

Commercial fleet launch 规划中

RaaS (Robot-as-a-Service) subscription model for high-value crops; 6× daily efficiency per unit targeted at scale.


Why CEA first?

Structured environments give robots their fastest path to reliability.