Autonomous Food Delivery Robot

Smart Environment Sensing


Fusion of LiDAR and camera sensor for exceptional smart environment sensing provides faster perception and safer delivery


UDU SLAM is a multi-sensor fusion SLAM algorithm independently developed by Pudu. It adopts a combination of various sensors such as Vision Camera, Lidar, IMU, RGB-D depth camera and ultrasonic radar. It can realize real-time centimeter-level self positioning and high-precision map construction. It can also run efficiently in complex and busy scenarios such as restaurants

Adjustable Tray

Metal Tray : 39 x 39 cm

Max. Load : 13 Kg. per tray

3D Obstacle Avoidance

Intel 3D Depth Camera sensor for safer delivery

New Lidar

New Generation LiDAR

More precise and faster SLAM Mapping

New Visual Sensor, More precise positioning

All-Aluminum FRAME 
Aviation-Grade Aluminum Alloy Frame
More stable structure

New suspension

Easily dealing with various bumps on the floor for stable operation

The new generation platform is equipped with top-speed LPDDR4 of 32Gb, dual-channel and 3733Mhz. This super configuration enables the robot to easily deal with the huge amount of data required in various complex environments, process more tasks in parallel and deal with more tedious scheduling.

Precise & POWERFUL environment sensing
Pudubot uses integrated data from multiple sensors; Lidar, vision camera with sensor pointing up, and 2 Intel RealSense Depth Cameras located in the lower part of the robot. This allows the robot to avoid collisions and adapt to the changing environment like in restaurants, which chairs and table legs moves, not to mention people moving around the space.

Powerful, smooth CONTROL with Linkage suspension
Pudubot has 6 wheels with well-designed independent suspension linkage system. This allows Pudubot to dealing with various bumps on the floor, with smooth and stable movement. The braking time is as short as 0.5s, preventing sudden collision. Pudubot can hold load up to max. 52 Kg.


Super-Easy to operate for waiters and customers

User-Friendly: With easy-to-use Graphic interface, Pudubot goes to work in just few taps on the screen. Multi-function and scenarios of operations for various types of uses e.g. Direct Delivery, Multiple-delivery, Auto-Cruise mode, Happy Birthday.

Customer-Friendly: With Adaptive screen and cute Voice (Support Thai Language) interface can guide customers to enjoy interacting with Pudubot in very short time learning.

The safety and obstacle avoidance performance have been greatly improved, and the probability of touching suspended objects or other blind spot obstacles has been greatly reduced, in addition to Lidar, multiple visual sensors are added to enable the robot to upgrade from 2D plane perception to 3D environment detection of the whole body.

Compared to traditional 2D radar, 3D stereo sensors can quickly stereo model the surrounding environment, which can greatly reduce the probability of security accidents.
Minimum height of object detection is 
Obstacle detection frequency up to 5400 times per minute 
Front obstacle detection range exceeds 10m.


PUDU SCHEDULER has adopted a decentralized, flexible ad hoc network communication scheme and algorithm architecture. Each robot can directly communicate with all the robots in the same place, and quickly manage and make decisions about traffic e.g. Making decision for robots to avoid going into the narrow isle at the same time, or a robot to wait at the junction to let another robot walk pass first, or robots queueing to wait for dishes one by one. This system makes multi-robot scenario more efficient and avoid jamming at busy spot





  • Machine size                     516 x 500 x 1288mm

  • Machine weight               PD1: 40kg PD6: 35kg

  • Machine material            ABS/aviation-grade aluminum alloy

  • Charging time                  4H

  • Battery lifetime               10~24H

  • Cruise speed                     0.5~1.2m/s adjustable

  • Single tray load                PD1: 10kgPD6: 13kg