The Robot That Brings Drink

Year: From 2018

Position: Invetor, Designer

Problem & Purpose

The difficulty with the chores especially the elderly and physically disabled people have and the demand of a robot to help them in their daily life is a huge problem today. The purpose of this project is to develop a robot to help a physically disabled, elder and sick people in where they live and to provide the possibility to command the robot to bring a drink via voice commands over a smartphone.

YouTube video of TASHIMASU

Open Source Code, Circuit Schematics, PCB Design, and 3D Models: CLICK HERE TO SEE

Literature Review

The robot this project is based on is the first instance of a robot that can stock the drinks, fill it to the glass and bring it to the user.

However there are some similar projects in the literature. In order to be talked about the similar robots, first the waiter robots should be mentioned like these three:

  • Akınsoft, ADA
  • Robot Sopo
  • Asus, Zenbo

Innovative Aspects

  • Aiming to take the work of drink bringing as the main task and to be developed for being capable to wider range of services.
  • Being able to fill the drinks to the glasses which are carried in bottles in the robot for a better and easier service for the users.
  • Using the complex ways like computer vision and face tracking rather than simple ways like line following to detect the location of the user.
  • Giving service with the voice commands via the voice assistant in the smartphones and tablets.
  • Being designed using low cost components and methods. (Costing 1250 TL )


Mechanics: Motion Mechanism

  • 4 pieces of 12-14v, 150 rpm, 42mm, 7,5watt, high torque DC motors Each connected to 4 independent wheels.
  • Wheels from common robotics kits, 3D printed connectors.
  • Tank steering

Mechanics: Body and Frame

  • 56x37x80 sizes, high enough for the access to the glass.
  • 3D designing software: Tinkercad.
  • Materials:
    • Main down body: 2mm sheet steel, low cost
    • Upper body: 5mm and 3mm MDF wood

Mechanics: Head

  • Adafruit, 2 axis “Pan-tilt” mechanism.
  • Horizontal and vertical movement.
  • Sg90 9g servos
  • Low cost and useful

Mechanics: Liquid Transfer System

  • Transfer of drinks to the glass from inside drink tanks.
  • 2 pieces of 12v DC peristaltic liquid pumps (Food grade).
  • Silicone fish tank hose for the transfer
    • Downside: Not appropriate for hot drinks

Mechanics: Load Cell

  • Is located just below the 3D printed platform which holds the glass, and prevents the spillages by measuring the weight of the drink in glass.
  • Is able to make measurements up to 1 kg.


Electronics: Battery

  • The energy consumption of the robot is very high due to components used (motors, pumps, raspberry pi …)
  • Two different power sources are tested and verified on this robot:
    • 11.1v 2800 mAh Li-Po Battery (Quick movements but short life)
    • 12v 7Ah Pb(lead) Accu (Slow movements but long life)

Electronics: Motor Drivers

  • For the main motion motors:
    • 2 pieces of L298N motor driver board
  • For the liquid pumps:
    • 1 piece of L298N motor driver board
  • Advantage:
    • Metal cooling piece, easy use, 5v regulator

Electronics: Microcontrollers (Arduino)

  • Raspberry Pi doesn’t have enough pins for the components used in the robot.
  • 3 pieces of Arduino Uno are used.
  • With the help of USB communication:
    • Arduino’s are controlled (Python serial communication)
    • And programmed. (Arduino IDE)
  • Also Arduino’s operating with 5v is an advantage.

Electronics: Raspberry Pi

  • Has enough of ram, (1GB RAM)
  • Fast and secure (1GHz)
  • Low energy consumption (100-400 mA)
  • Small sized and cheap ( Credit-card sized, $35)

A Raspberry Pi 3 Model B is used for the robot which has the specifications above.

Electronics: Load Cell Amplifier Module

  • In order to read the low voltage analog data coming from the load cell:
    • Analog data is required to be amplified
    • and digitalized.
  • For this “HX711” module ,which is an ADC (analog-digital converter) with an amplifier function, is used.

Electronics: Distance Sensors

  • These sensors are placed to four sides of robot to avoid collisions in autonomous mode.
  • Hc-sr04 ultrasonic distance sensors are used.

Electronics: Homemade Motherboard

Because of the reason that the breadboards are not so reliable to use a “motherboard” for the robot is produced using a perfboard with THT(Trough-hole) technology.

  • Provides electrical connection between all components of the robot
  • Avoids the cable compexity
  • 6×13 cm sizes
  • PCB design is planned

Electronics: LED Strip Lighting

In order to make the robot appealing to eye at the same time

  • Around 30 RGB LED’s are used.
  • LED’s are in the form of a strip.
  • Can be controlled with only 2 pins on Arduino.
  • Can visually show different animations according to the motion of the robot.

Electronics: Camera

  • A Raspberry Pi camera is used for the robot to do computer vision and face tracking applications.
    • 3 grams
    • 2×2 cm sizes
    • 5mp resolution
    • Low cost
  • A case was 3D printed for the camera.

Electronics: Speaker

The robot can speak Turkish in order to greet, warn and entertain people in autonomous mode. And for the robot to speak a simple, battery included, 3.5mm jack bluetooth speaker is used.

Electronics: 2×16 LCD Screen

  • Tashimasu with 2×16 LCD character screen,
  • Is able to deliver various warnings and messages to the people around.
    • Can also be used as the “mouth” of the robot.

Electronics: Powerbank (Portable Power Source)

  • Raspberry Pi requires a stable 5V power.
  • Regulators on the L298N can’t deliver enough power.
    • For these reasons Raspberry Pi 3 is powered with the help of a powerbank in the Tashimasu robot.
  • 4000 mAh capacity
  • 5v 2A output capability

Circuit Schematic


Software: Raspberry Pi and Python

  • A Python Virtual Environment is created.
  • OpenCV library is used for computer vision. “face2.py”
  • “software2.py” (main script) is runned whenever the robot is powered up. And robot switches to the autonomous mode.
  • Or this program can be runned through SSH in order to control the robot manually.

Software: Google Assistant, IFTTT and Blynk

  • Initially the voice commands are captured from the smartphone using Google Asistant.
  • Then they are transferred to the Blynk server via webhooks and with the help of IFTTT service.
  • After that the Blynk server delivers the command to the NodeMCU board which has the Blynk arduino software flashed.

Software: NodeMCU Board (ESP8266

  • NodeMCU, is an iot board connected to the Internet via wifi.
  • In this project it is programmed with Arduino IDE.
  • The commands taken from the Blynk are transferred to the NodeMCU
  • And the NodeMCU delivers the received commands to Raspberry Pi.

General Mindmap

Software: Arduino-A / Motors

  • The commands used for the motion of the robot are; “forward”, “reverse”, “turn left”, “turn right”, “pivot left”, “pivot right” ve “stop” …
  • The control of the liquid pumps,
  • Servo control,
  • To receive the data from HX71

Are done by this Arduino.

Software: Arduino-B / Sensors

  • Transfers the data coming from the Hc-Sr04 distance sensors on the 4 sides of the robot to the Raspberry Pi via USB. This it avoids collisions.
  • Provides the control of the RGB LED strip lighting around the robot with the help of “Adafruit_WS2801” library.

Software: Arduino-C / LCD Screen

  • Controls 2×16 character LCD screen.
  • Uses Arduino’s own “LiquidCrystal” library.
  • Various texts and animations are shown on this LCD.

Operation Algorithm Mindmap


  • Various methods are investigated for the robot to receive the voice commands from the Google Assistant; with the use of IFTTT, Blynk servers and NodeMCU board, an integrative solution is formed by combining different systems.
  • The inference of that “a better system is required to be used because of the reason that the liquid pumps are not able to fill up the glass faster than 120 seconds is” made.
  • Also it is proved that “in order to reach to the solutions in engineering it is essential to connect different dots” by using various programming languages and multiple controllers in an unusual way.

Conclusion and Discussion

  • Consequently it is attained that the robot can come besides the users and serve the drink by filling the glass from two types of drink. And this is done by making the robot run control algorithms which are based on preprogrammed commands after making the robot receive the voice commands given from the personal assistants on mobile devices
    • In this way, the liquid need of physically handicapped and elder is fulfilled without requiring them to walk via voice commands and thus their life is made easier.
    • In addition to this the voice commands can be given to the robot from anywhere on the world.
    • The computer vision on the robot is sufficiently advanced to be able to determine the location of the users however it is possible to work on it furthermore.


  • Motion mechanism can be improved. (For example legged or tracked…)
  • Glass can be placed to a higher place in the new body.
  • Pumps with with a higher flow and a battery with a higher capacity can be used.
  • Transiting to the pruduction can be madde easier by using a PCB printed circuit.
  • Haarcascade of the OpenCV can be improved.
  • In order to make the robot operate in the rooms where more than one people are, a new model can be trained using the OpenCV and tensorflow and also with the use of GoogleCloudVisionAPI somewhat a “FaceID” technology can be developed. Thus the faces of users can be matched with their google accounts.
  • Also with the conditions like the absence of the mobile phone the command of the user can be received directly from the microphones placed to different spots of the robot. In this way, the command, the location of the user and the identification of the user can be detected.
  • By training new computer vision models using OpenCV and tensorflow, the robot can be made to recognize it’s surroundings better and interact with them too. (For example opening a door, going over the carpet…)
  • Or the robot can be made to recognize it’s surrounding objects using GoogleCloudVisionAPI.
  • Also with these and further possible improvements, robot can be made to help with the chores in the house in addition to bringing a drink in order to ease our lives moreover. (Cleaning, roundup, dishes, food, preparing the table for the dinner, laundry…)

Feasibility and Common Effect

  • Robot can be put into production using of practical and industrial solutions with the help of the robot’s simple structure and it’s common components.
  • Also it is possible to produce this robot at home too. Therefore an open source developer community can be formed via sharing it’s all software and design files for free.
  • It can be expressed that the robot has a very common effect because the places that it is developed to be used are not only houses but cafes, restaurants, hotels and schools and many more… Which are the places where elder people spend their time a lot.

Poster of The TASHIMASU Which is Used in Competitions: CLICK HERE

Awards This Project Has Received

Championship at INSPO 2018 (Engineering)

Okyanus Colleges / First place in the International Science Project Olympiad held in Okyanus Colleges

6th Place at Information and Technology Competition in İZMİT

BTV (Information and Technology Foundation) / 6th place at Information and Technology Competition held by the Information and Technology Foundation and the municipality of Izmit with the approval of the Republic of Turkey Ministry of National Education

Recent Progress

PCB design was made and ordered from JLCPCB.

The problematic wheels have been replaced.

Presentation Deck

Project Development with a Global Household Appliances Company

This project was being developed in partnership with a major household appliances company both in Turkey and worldwide, in this way this project was planned to move to a more professional step such as mass production.

Photographs of TASHIMASU


  • [1]- Date of access: 14.03.2018, https://www.raspberrypi.org/
  • [2]- EDate of access: 14.03.2018, https://www.python.org/
  • [3]- Bozkurt, S (2012). Servo Motor Nedir? ve Servo Kontrolü, Date of access: 14.03.2018, http://www.teknokoliker.com/2012/07/servo-motorlar-ve-kullanm.html
  • [4]- SG90 Analog, Date of access: 14.03.2018, http://www.towerpro.com.tw/product/sg90-analog/
  • [5]- Raspberry Pi 3 Model B, Date of access: 14.03.2018, https://www.raspberrypi.org/products/raspberry-pi-3-model-b/
  • [6]- Date of access: 14.03.2018, http://simplecv.org/
  • [7]- Admin (2011). Redüktörlü Motorlar, Date of access:09.04.2018, http://www.elektrikrehberiniz.com/elektrik-motorlari/reduktorlu-motorlar-578/
  • [8]- Peristaltik Hortum Pompalar, Date of access:09.04.2018, http://www.leonteknik.com/pompa/peristaltik-pompa
  • [9]- Mançar, B. (2014). IFTTT İle Hayatınızı Otomatiğe Bağlayın, Date of access:09.04.2018, http://www.webmasto.com/ifttt-nedir-nasil-kullanilir
  • [10]- NodeMCU Nedir?, Date of access: 9.08.2018, https://www.bilisimhocasi.com/nodemcu-nedir
  • [11]- JavaScript, Date of access:11.04.2018, https://www.javascript.com/
  • [12]- Meet Sopo, the robot that brings you drinks, Date of access:11.04.2018, https://www.youtube.com/watch?v=uKv1GCzzVZc
  • [13]- Ramazan, M.(2017). Garson Robotlar, Date of access:11.04.2018, http://roboturka.com/robotik/garson-robotlar/
  • [14]- Türkiye’nin ilk robot garsonu işbaşı yaptı, Date of access:11.04.2018, http://www.milliyet.com.tr/turkiye-nin-ilk-robot-garsonu-gundem-2035026/
  • [15]- Ulukan, G. (2018). Alexa bağlantılı yeni robot, içeceğinizi ayağınıza getiriyor, Date of access:11.04.2018, https://webrazzi.com/2018/03/27/alexa-baglantili-yeni-robot-iceceginizi-ayaginiza-getiriyor/
  • [16]- Köse, A. (2017). Hizmetçi Robot Asus Zenbo Tanıtıldı ve Fiyatı Açıklandı!, Date of access:11.04.2018, http://www.webtekno.com/zimbirtilar/asus-ev-hizmetcisi-robotu-zenbo-yu-tanitti-ve-fiyatini-acikladi-h17439.html

Leave a Reply

Your email address will not be published. Required fields are marked *