Multi-Disciplinary Engineering Design Conference

Kate Gleason College of Engineering

Rochester Institute of Technology
Rochester, New York 14623

Project Number: P07403

GARRI PROCESSING
Alissa Albert/Industrial Engineer

Anne Froustet/Industrial Engineer

Coleman Glasgow/Mechanical Engineer

Joseph Lawson/Mechanical Engineer

ABSTRACT
The production of Garri, the staple carbohydrate of
West African nations, requires a time consuming and
labor intensive process in order to transform the raw
material, cassava roots, into the finished product. A
garri processing machine was chosen as a focus
project for RITs Multidisciplinary Senior Design
program in the sustainable systems track. The first
two sub-processes of producing garri were identified
as being particularly difficult and time consuming, so
a machine designed to automate each of these
processes was developed. A peeling machine was
successfully developed to meet or exceed all needed
specifications. A grating machine was also
developed, however, further design iterations are still
needed to meet all requirements.
INTRODUCTION
The Garri Processing Project was suggested by RIT
alumnus, Eze Kamanu. Eze grew up in West Africa
and has been involved first hand with cassava/garri
processing. Eze identified current production
methods as being time-consuming and laborintensive.
Garri is the staple carbohydrate of West African
nations. Garri is a fine grain and can be made into
breads, crackers and pancakes. Garri is produced
from cassava, tuberous roots with thick skin and
dense flesh. An average family of six produces
approximately 50 kg of garri per month; this quantity
requires about 400 cassava roots. Traditional
processing includes six steps to achieving the final
product of garri: peeling, washing, grating, drying,
sifting, and frying.

The goal of the project was to design and

manufacture a garri processing machine affordable to
lower class families in Nigeria. The device would
improve upon one or more of the inefficient
processes related to the processing of garri. A
discussion with the customer provided details
regarding methods and time estimates for each step.
The two steps of peeling and grating chosen for
concentration were selected based on the given
information and the teams ability to make a
significant improvement. The peeling process
traditionally takes three to four people, three to four
laborious hours of peeling the monthly supply,
utilizing only a standard kitchen knife. The current
grating method involves renting a machine that takes
about one half hour to one hour to perform the entire
grating for a months supply. While the grating
operation itself is relatively quick, families can wait
up to a full day for the machine to arrive at their
home.
The primary objective of the project was to develop a
device capable of peeling and grating cassava. This
machine should reduce processing time and labor
required for peeling. The product should also
eliminate the reliance on a shared grating machine for
West African families. This should save time and
money associated with renting the machine. Finally,
the product should be durable, easy and safe to use,
and affordable to lower class families.
The expected benefit of the Garri Processing Senior
Design Project is to improve the quality of life for
approximately 8-million Nigerian families by
providing more efficient, economical methods for
garri processing.

2007 Rochester Institute of Technology

Proceedings of the Multi-Disciplinary Engineering Design Conference

CONCEPT GENERATION
The concept generation for the group began with a
benchmarking analysis of products available on the
market that could be applied to processing cassavas.
Machines currently available in West Africa and
those in the United States were compared to generate
ideas on how to effectively peel and grate the roots.
The most popular grating machine currently used in
West Africa can be rented, posing an accessibility
issue as customers cannot use it whenever they want.
Taking into consideration the customers the group
was looking to market to, it was decided that the best
approach to designing machinery for such processing
would involve manual power to keep cost down and
provide allowance for using the products in any
setting. The two most effective products evaluated
for benchmarking were an apple peeler and a food
mill.
Initial concepts generated ranged from one large
machine that would combine peeling and grating in
one unit, driving cassavas on rollers to more basic
concepts like abrasive rollers with teeth rubbing the
skin off. All concepts were scored and ranked in a
matrix to determine which would be chosen for the
initial concept. Concept scoring involved comparing
the critical features and components of the conceptual
designs with those of the benchmarked products. The
concept initially selected involved both peeling and
grating, incorporating both a blade similar to that on
an apple peeler and a fine grating disc. The cassavas
would be fed through an opening in a box while the
user cranks a handle to turn the blade arm and disc.
The cassava would be peeled and then grated into a
fine grain and deposited into a receptacle at the
bottom of the box.
After the initial concept was presented, professors
and students commented on improvements and
considerations for the design including: adjusting the
grating disc so that it would sit off-center of the
central axis for effective grating, and altering the
drive of the blade and disc to motorized power rather
than manual. Upon reflection of the critique of the
original concept, the decision was made to simplify
the design by breaking up the peeling and grating into
two separate, smaller machines. This would make
the goal of creating products within the set cost range
more reachable.
The change in direction of the project required the
original benchmarking analysis to be revisited, with
smaller machines. The apple peeler idea was
maintained as a basis for what to use to remove the
skin of the cassava. Further brainstorming and
research led to the development of using a rotating
cone with teeth for grating. Similar products are
presently available, but the initial concept required

Page 2

modification to meet the constraints of the project.

The products on the market now work for smaller
fruits and vegetables like apples and potatoes. The
length of the cassava caused concern and presented
an opportunity for improvement for the apple peeler
to accommodate larger pieces of roots. By
purchasing only the cone for the grating device, more
flexibility could be incorporated into the design of
the apparatus. This allows for appropriate
development of safety features.
MANUFACTURING
A modified version of the apple peeler design found
in the benchmarking process was adapted to meet the
needs of peeling cassava. The existing peeler design
consisted of a suction base so that it can be attached
to a relatively smooth surface. This was the first
point of the peeler to be modified so that the device
would be capable of being affixed to a multitude of
table surfaces and thicknesses. The second design
modification came after realizing that the cassava
could potentially out weigh the ability for the threeprongs to support the load; for this, a spring support
with a roller bead was devised to support the cassava.
This provides support for cassavas while being driven
and rotated into the peeling blade. To achieve both
of these feats, a c-clamp was cut and welded out of
1 steel bar stock and attached to the bottom of the
peeler using the pre-existing threaded holes from the
suction base. The support system was created out of
this same bar stock by heating and bending two
upright posts which the springs and roller were strung
between. This support structure was welded to the
end of the c-clamp to maintain a flush table-to-peeler
tolerance and to reduce the need to attach it to the
peeler separately from the c-clamp.
A structure to house and rotate the cone for the
grating portion of the project was also fabricated. A
method for rotating the cone was accomplished by
creating a drive mechanism with six teeth to turn the
cone on a central axis. The central axis, for which
the cone rotates on, consists of a carriage bolt driven
by a handle made of glass-filled nylon. The whole
assembly is supported by a pipe, in which, the
carriage bolt rotates. From here, a base was
constructed out of steel bar stock. The base includes a
c-clamp to attach the device to different work
surfaces. The base is also designed to provide enough
clearance between the cone and the work surface to
allow for the cassava to exit the cone without backing
up. The rotating blade of the grating machine posed
a safety risk towards users so additional
manufacturing designs became necessary. The shield
and chute designs chosen both provide protection
from the moving grater cone, and assist in directing
the cassava into the moving grater. The chute adds
additional safety in that it is 6 long and narrow

Paper Number P07403

Proceedings of the KGCOE Multi-Disciplinary Engineering Design Conference

enough that a person could not reach their hand and
fingers into the moving parts. For prototype
purposes, the shield consists of a steel can and the
chute is constructed out of a thick cardstock tube. A
grinder was used to cut the can so as to accommodate
the cardboard chute. The two were adhered together
and then the chute was trimmed to allow the cone to
fit in the new safety housing. The shield and chute
assembly are attached to the rest of the machine using
machine screws.
The material chosen for these two designs were for
prototype purposes only. The materials were easy to
manipulate and readily accessible. For the intent of
mass production, other materials such as plastics for
the shield and powdered metals for both the grater
and peeler would be preferential.
TESTING
Testing was necessary to determine the success of the
selected peeling and grating designs. The time and
effort to perform these tasks are the most critical
improvements to be made. Testing methods were
designed to identify the strong and weak points of
each design and confirm that product specifications
are met.
Cycle Time:
Time trials were performed to obtain the average
cycle time for both the peeling and grating devices.
Due to the inability to obtain cassava in the Rochester
area, all time trial tests were performed with a yucca
root, which is very similar in size, shape, and skin
and flesh texture. Yucca samples were prepared in
four inch sections. The diameter of each end was
measured and recorded along with the mass of each
sample. Each sample was peeled, weighed again and
then grated while times were recorded for the peeling
and grating portions of the testing. The average
length of a cassava is actually six inches long making
it necessary to weight the times recorded for both
peeling and grating. The time recorded to grate each
yucca sample was multiplied by a factor of 3/2 in
order to account for this variance. This weighting
factor is appropriate because there is very little, if
any, time required to change over and begin grating a
second cassava. When peeling, two time
measurements were recorded: the average time
required to load and unload the peeling apparatus,
and the average time required to actually peel each
yucca sample. To obtain an average cycle time for
the peeler the peeling time was multiplied by a factor
of 3/2 and added to the loading time because the
loading time has a negligible scaling factor compared
to the actual time to peel.

Page 3

Percent Mass Removal:

Using the mass of each sample before and after the
peeling process, the percent mass removal was
calculated. A similar test was performed in which a
yucca was peeled manually using a knife to
determine the percent mass removal using traditional
methods. There was no time limit on the manual
peeling to insure safe and appropriate usage of the
knife.
Usability Testing:
A usability test was performed in order to determine
the feasibility of these products being used in place of
the traditional methods. A request was made to the
RIT Internal Review Board for Human Subject
Testing and all necessary paperwork was filled out so
that human testing could be conducted on RIT
grounds. A group of 20 random participants were
presented both the peeler and grater and given a
portion of a yam. They were told what each machine
did but were not given instruction on how to use
either. When the user was finished, they were asked
to fill out an exit survey which was used to determine
whether the machines were intuitive and what their
likelihood of using them would be.
RESULTS
Completion of usability studies and time trials allows
for an analysis of the results, comparing the test
findings to the product specifications. The
specifications under investigation in the user trials
included intuitive product use, ease of use as a
function of the force required to operate the devices,
and the users product preference (the marketability
of the product). The time trials, conducted by the
engineering design team, investigated the average
time to peel and grate cassavas.
User Surveys:
The results of the user studies were favorable. For
both the peeler and the grater, the target specification
for acceptance was 100% of users, with a marginal
value of 75%. While the team was willing to accept
75% for product preference and ease of use, the goal
was to achieve as high a value as possible,
understanding that some users may favor a product
they are more comfortable or familiar with.
The peeler studies (figure 1) found that 95% of users
could operate the peeler without instruction and
100% of users were comfortable with the amount of
force required to peel a yam. When users were asked
about their product preference, 90% of users would
choose the peeler over a standard kitchen knife, 5%
of users would prefer to use a knife, and 5% were
either undecided or had never peeled anything with a
knife and thus had no basis for comparison. These

Copyright 2007 by Rochester Institute of Technology

Proceedings of the Multi-Disciplinary Engineering Design Conference

findings fall well within the specified range of
acceptance.
Peeler Usability Ratings
100%

Target

95%
90%
85%
80%
75%

Marginal Value

70%

Intuitive Use

Force Required

Figure 1. User survey results of peeler

The user trials conducted on the grater (figure 2)

resulted in a varied response from users. While all of
the users found the use of the grater to be rather
intuitive, the users product preferences were quite
varied. The amount of force required to operate the
grater appeared to be a limiting factor; 65% of the

Mass Removal Comparison:

The cassava peeler resulted in an average mass
removal of 12.7% of the original mass. When peeled
with a knife, the average mass removed from the
cassava was 15.9%. This is a pass/fail test and the
peeler passes since the amount of skin and flesh
removed is less than that of traditional methods,
using a knife.
Cycle Time:
For peeling, the marginal and target specifications
were 30 and 15 seconds per cassava, respectively.
Based on the production of 100 cassavas per week
per family, these specifications would yield a product
that was capable of peeling 100 cassavas in 25-50
minutes (see figure 3). The average time to peel a
cassava, computed from the trial data, was 22.75
seconds or approximately 38 minutes spent per week
peeling cassavas (see figure 4). With traditional
methods requiring between two and a quarter and
four man-hours per week to peel cassavas, this
method will reduce processing time by over 70%.

Grater Usability Ratings

100%

Page 4

Peeling Time Trials

30

Target

Marginal Value

Seconds/6" Cassava

95%
25

90%
85%
80%
75%

22.75

20

Marginal Value

15

70%

Target
Average Test
Time

65%
60%

Intuitive Use

Force Required

10

Figure 2. User survey results of grater

Figure 3. Peeler time trials

users were comfortable with the amount of force used

and believed they would be able to grate 50 cassavas
or yams with this device. Only 50% of users
indicated that they would choose this grater over a
cheese grater, 30% would choose the cheese grater,
and 20% were either undecided or had never used a
cheese grater and had no comparison. It is believed
that there is a strong correlation between product
preference and the amount of force required to use
the grater. Many of the user comments on the surveys
indicated force as a problem; grating 50
cassava/yams could be tiring or require multiple
breaks. If the force is reduced, it is likely the product
would be more widely accepted. The force and
product preference ratings fall below the 75%
acceptance and resulted in brainstorming some
modifications to the design and recommendations for
future work.

Grater Time Trials

Seconds/6" Cassava

130
120
110
100
90
80
70
60
50
40
30
20
10
0

Average Test
Time

125.1

Marginal Value
Target Value

Figure 4. Grater time trials

The product specifications for the grater require a

cassava to be grated in, at most, 60 seconds, with a
target of 30 seconds. These specifications would
ensure that the product could grater a weeks supply
of cassava in a maximum of 1 hour and 40 minutes.
The time trials produced an average grating time of

Paper Number P07403

Proceedings of the KGCOE Multi-Disciplinary Engineering Design Conference

125.1 seconds per cassava. This is twice the marginal
acceptable value and would require over 3 hours and
20 minutes to grate 100 cassavas.
The user studies and time trials conducted for the
peeler indicate that the device both performs to
product specifications and has user preference and
acceptance. The grater, on the other hand, requires
some modifications to increase user acceptance and
presently falls outside of the product specification
range for acceptable grating time.
CONCLUSION
The peeling machine significantly reduces the time
associated with peeling cassava for garri production

Page 5

while making the entire process more bearable for the

user. The current design of the grater proved
inefficient in reducing either the time or effort
required. Future work needs to be performed in order
to stabilize the assembly and make the handle easier
to turn to reduce the required force. This concept
shows potential though, that with a few more design
iterations it could become a feasible design.
ACKNOWLEDGEMENTS
The team wishes to thank Dr. Taylor, Maggie
Masella and Eze Kamanu as well as Dr. Hensel and
the Mechanical Engineering Department, for their
support and sponsorship

Copyright 2007 by Rochester Institute of Technology