Full Paper ICAS2017 Submitted

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/318695151

IoT based Coffee quality monitoring and processing system in Rwanda

Conference Paper · May 2017


DOI: 10.1109/ICASI.2017.7988106

CITATIONS READS
2 3,031

3 authors:

Joachim Rutayisire Sandor Markon

1 PUBLICATION   2 CITATIONS   
Kobe Institute of Computing
110 PUBLICATIONS   1,015 CITATIONS   
SEE PROFILE
SEE PROFILE

Ndacyayisaba Raymond
University of Hawaiʻi at Mānoa
1 PUBLICATION   2 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Multisensory interaction with floating images View project

BongoFaster Mobile App for BRT in Dar es Salaam Tanzania View project

All content following this page was uploaded by Sandor Markon on 22 September 2018.

The user has requested enhancement of the downloaded file.


IoT Based Coffee Quality Monitoring and Processing System in Rwanda
Joachim Rutayisire1, Sandor Markon2 and Ndacyayisaba Raymond3
1-2 Department of Information Technology , 3Department of Electronic and Telecommunications
1-2 Kobe Institute of Computing – KIC, 3Tumba college of Technology - TCT
1-2 Kobe City, Japan , 3 Rulindo city, Rwanda
[email protected] 1
[email protected] 2
[email protected] 3

Abstract quality of Rwandan coffee at the international market, The


biggest problems are to control the standard conditions of
Coffee is Rwanda's main source of income, thus improving coffee treatment to return high quality product due to the lack
processing and monitoring has a large potential economical of controlled stages of processing, Fermentation and drying.
impact, we are introducing low-cost sensor Technology, by whereas, they still use traditional means of sun dying and
utilizing embedded system to monitor coffee washing station fermentation method which can damage the quality of cherries
processing method and store centers, the system will be and eventually the quality of coffee when is not well
reporting the status of PH, moisture, temperature and humidity controlled [4]
as basic information to keep the standard quality of coffee,
continually monitor these stages and alerting right time to We believe that similar technologies could be beneficent in
move from one stage to another to meet requirement of export coffee growing areas not just in Rwanda but in other countries
market. In this paper we describe the background of coffee with similar conditions of coffee processing [1].
quality major issues through processing and monitoring, our
proposed e-Kawa solution and preliminary experimental II. Current Solution
results obtained in lab testing. We believe that similar
technologies could be beneficent in local coffee processing Coffee processing in Rwanda is using washing station located
anywhere else with similar conditions. in rural areas. Processing of coffee beans is conducted in four
stages; removing cover of coffee, washing and ferment for a
Keywords: IoT, fermentation,washing station, pH, e-Kawa while inside water tank and drying coffee to be packaged to
the market. There is no current method to track realtime to
I. Introduction remove coffee beans from one stage to another stage of pro-
cessing, where each stage affects the quality of coffee beans at
Coffee is Rwanda’s premier export and source of foreign cur- the end of process and has some factors to consider in shifting
rency with the potential to lift farmers in rural areas and pro- stages (washing, fermentation, drying and storing) as main
cessing companies out of poverty. Coffee growing in Rwanda stages to keep standard quality required by international mar-
started in 1930, when the Belgian colonial government re- ket [3].
quired Rwandan farmers to plant coffee trees, while setting
price restrictions and high export taxes, and controlling which III. General Objective
firms could purchase coffee [2].
In the coffee value chain, According to statistics from the Na- The main aim of this study after analyzing factors that influ-
tional Agricultural Exports Board (NAEB), in 2014 -2015 encing quality while washing coffee, drying and stocking is to
Rwanda fetched $64.02 million from coffee exports, up from develop low-cost system that can detect, visualize data and
$54.9 million in 2013 from an annual coffee production be- alert next stages to be taken.
tween 18000T to 21000T (NAEB 2015).per year. we will focus on different challenges, some of them were
The recent NAEB-IGC coffee conference in Kigali (February, raised:
2015) has highlighted two key priorities for the industry
Improve access to inputs and extension services for farmers • Stability of coffee quality required by international market
and established a regulatory framework to improve the • Human errors due to the duration of fermentation, drying
relationship between farmers and coffee washing stations and process and related factors as a major key to result seed to
ensure a reasonable share of the international prices reaches cup quality.
the farmers. Washing station has been constructed to improve • Lack of standard infrastructure to maintain weather and
the quality of coffee, coffee processing companies are keep the quality from processing to the market
innovating current washing stations, washing tanks, irrigation
and stocks to maintain coffee quality (NAEB 2014). IV. Method
Significant public investments were made in infrastructure,
setting up cupping labs to develop technical expertise, and This research was conducted under coffee washing stations
applying for certification programs such as Starbucks Coffee in Rwanda, we have conducted interview with WestRock
and Farmer Equity (CAFE). Certification harvest-handling has company technicians and washing stations coordinators on
been identified as one of the major factors threatening the
current coffee processing and challenges they are facing in and sensors that could be used for assisting coffee growers to
daily processing. boost the quality of coffee beans [10].

Coffee washing stations are located in rural area where farm-


ers (co-op) randomly supply coffee beans to be processed on
field by different exporters companies V. Related Work

World coffee trade-Grading and classification is usually based


on some number of criteria; Altitude or region, Preparation
method used for processing either washed or natural process,
screen size of bean, sometimes also color, Number of imper-
fections as the main key to result cup quality flavor.

Mechanical coffee driers

Some developed countries have developed big industries with


Fig 1. Fermentation process mechanical drying the beans are heated by the passage of hot
air which also carries the moisture away. Temperatures must
This process is fermentation, a few hours after the cherries be monitored during natural and artificial drying. Coffee
are harvested, the pulp is removed from the cherries. The temperature should not exceed 40°C for parchment and 45°C
beans are then washed in a process that involves cycles of for cherries. It is often thought that overheating can only occur
fermentation which takes about 24 to 40 hours inside water in mechanical dryers [5].
tanks, the total time of hours depends on numerous factors
such us water temperature and humidity, degree of cherry There are mainly 2 types of dryers, static and revolving. In
maturation, altitude. Higher altitude due to cooler temperature revolving dryers, there are tray dryers with stirrer, vertical
require more time in fermentation than the ones at sea level. dryers and rotary dryers, cascade driers, column driers, and
The mucilage is made up of pectin materials, the fresh flex driers. In all the cases woods, coffee husk, other solid
mucilage has a pH of about 6.5 after harvesting, during fuel, fuel oils, diesel, gases are used as the main fuel or energy
fermentation this falls rapidly to a minimum of about 4.1 to sources. In case of mechanical dryers drying time varies from
4.3 by reducing sugars including glucose and fructose (30%), 20-60 hours according to the type of driers used .The main
non-reducing sugars such as sucrose (20%), cellulose and ash disadvantage of the mechanical dryers are that the drying is
(17%) (Wrigley, 455). Coffee usually tend to fetch a higher not uniform[7].
price due to the increased amount of labor associated with
it[2]. Fermentation PH level analyzer
Studies showed that monitoring the pH by taking measures of
the pH of the fermentation solution, farmers receive accurate
readings of how far along the beans are and when to remove
the beans from the solution, thereby delivering a higher output
of quality beans Dobson and Hughes provided the pH device
to several dozen farmers with testing electronics and the
strainers protecting the pH probes [2].

VI. Proposed System Architecture

Fig 2. Drying process

Drying under natural process can take 15 up to 20 days.


The dried cherries are then hulled by hand, or machine to
remove the dried out pulp . This is the way coffee has been
processed from centuries.
Freshly pulped coffee has a moisture content (mc) of about
55%, until fully dry(12 -11 % mc)
Research carried out on the harvesting process and preparing
coffee beans revealed that small farmers lack data on a crucial
step that affects quality. We propose introducing low-cost and
open source embedded system consists of micro-controller

Fig 3. Proposed System Architecture


The sample data collected from the fermentation and
The proposed system architecture is composed with various different liquids are showing level of PH collected by e-Kawa
types of low cost environmental sensors (Temperature, system sending data to Ubidots cloud platform (Fig 5)
humidity, Coffee beans moisture and PH sensor) connected to
wifi micro-controller. In this research, we have used
WROOM 02
and NodeMCU 8266 integrated with wifi module that can
transmit data to the nearby access point and upload data to
cloud platform (Ubidots ).

VII. The Proposed System

We have tested our sensors networks, the results proved that


this research can be used to collect data in different location
of washing station and also alert for next stage to be taken,
even-though we need to maintain our sensors to provide accu-
rate data[2]. Fig 6. PH level statistics

TABLE 1: COFFEE PROCESSING LEVELS

ALERTING COFFEE PROCESSING STAGES

FERMENTATION FRESH
DRY COFFEE CHERRY
CHERRY

Start End Start End


Hours Days
Level Level Moi Level Level
sture
PH cont
Level ent
4.1 24 11 15
6.5 ~ ~ (%) 65 ~ ~
Fig 4. Proposed System
4.3 40 12 20

In the field, coffee factory workers will place the Ph sensor


into the fermentation tanks for coffee cherries, where the sen-
sors autonomously monitor the stage of the fermentation
process. Workers will receive alerts from the system when an Many study and field testing statistics have been conducted
intervention is required. This will allow reducing the frequen- since and the results proved the standards of quality coffee
cy of human monitoring, and remove the possibility of human based on level of PH and moisture content.
error causing processing failures.
IX. Recommendation

VIII. Evaluation Strategy We are recommending coffee growing countries especially


developing countries who are still using natural processing to
In this paper, the evaluation method of the proposed IoT adopt low-cost IoT based coffee processing and monitoring
Coffee quality processing and monitoring system will be based system as results prove that can serve to improve coffee quali-
on usability of the system, the accuracy of data collected by ty in this field and related environment
sensors and the validity of the system.
X. Conclusions

We have verified that our e-Kawa system can read accurate


data from sensors and alert next process in coffee processing
stations,We are still improving our work to make it more user
friendly, affordable to the society and fit washing stations
environment in Rwanda. We are also working with different
stakeholders of coffee processing to improve production by
utilizing the same technology in similar processing and add
more futures to separate defected beans from good ones using
embedded system.

Fig 5 Fermentation Level Prototype


REFERENCES

[1] L. Tercanlioglu, “Postgraduate Students ’ use of Reading


Strategies in L1 and ESL Contexts  : Links to Success 1,”
vol. 5, no. 4, pp. 562–570, 2004.
[2] Langnau, Leslie. "3D Printed Ph Probe Measures
Fermentation Acidity For The Best Flavor Coffee". Design
World. N.p., 2017. Web. 25 Feb. 2017.
[3] "Coffees". Bourbon Coffee. N.p., 2017. Web. 16 Mar.
2017.
[4] "New Rwanda Light Roast Coffee! - Tiny Footprint
Coffee". Tiny Footprint Coffee. N.p., 2017. Web. 16 Mar.
2017.
[5] "Mechanical Dryer | The Coffee Shrub". Coffeeshrub.com.
N.p., 2017. Web. 16 Mar. 2017.
[6] Sivamani, Saraswathi, Namjin Bae, and Yongyun Cho. "A
Smart Service Model Based On Ubiquitous Sensor
Networks Using Vertical Farm Ontology". International
Journal of Distributed Sensor Networks 9.12 (2013):
161495. Web.
[7] Richardin, Pascale. "Https://Journals.Uair.Arizona.Edu/
Index.Php/Radiocarbon/Article/View/16146". Radiocarbon
55.3–4 (2013): n. pag. Web.
[8] Selmar, D. et al. "Germination Of Coffee Seeds And Its
Significance For Coffee Quality". Plant Biology 8.2
(2006): 260-264. Web.
[9] "British Coffee Association - Processing".
Britishcoffeeassociation.org. N.p., 2017. Web. 16 Mar.
2017.
[10] Levi, John F. S. "Robert H. Bates And Michael E. Lofchie
(Eds.) Agricultural Development In Africa: Issues Of
Public Policy, New York: Praeger, 1980.

View publication stats

You might also like