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

net/publication/315754361

Assessing IOT Projects in University Education - A Framework for Problem-

Based Learning

Conference Paper · May 2017

DOI: 10.1109/ICSE-SEET.2017.6

CITATIONS READS
17 2,087

5 authors, including:

Hanna Mäenpää Samu Varjonen

University of Helsinki University of Helsinki
14 PUBLICATIONS   175 CITATIONS    15 PUBLICATIONS   48 CITATIONS   

SEE PROFILE SEE PROFILE

Sasu Tarkoma
University of Helsinki
284 PUBLICATIONS   3,756 CITATIONS   

SEE PROFILE

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

Cooperation-Aware Software and Creative Self-Adaptivity (CACS) View project

Special Issue, IEEE Pervasive Computing View project

All content following this page was uploaded by Hanna Mäenpää on 07 November 2017.

The user has requested enhancement of the downloaded file.

 Blending Problem- and Project-Based Learning in Internet
of Things Education: Case Greenhouse Maintenance

Hanna Mäenpää Sasu Tarkoma Samu Varjonen Arto Vihavainen

Department of Computer Science

University of Helsinki
Helsinki, Finland
<firstname.lastname>@cs.helsinki.fi

ABSTRACT plines, software engineers play a central role in solving prob-

This article describes an experimental course where students lems, including environmental challenges [10]. This work is
developed Internet of Things device prototypes to improve done in teams where experts from various fields come to-
the upkeep of an urban rooftop greenhouse. With the help gether to share their knowledge.
of a problem-based learning approach, students were first In this article, we describe an experimental blended prob-
familiarized with their new learning environment and en- lem- and project-based learning course, where students de-
couraged to find issues that could be improved as a mean- velop solutions to support taking care of a greenhouse. The
ingful personal learning experience. A project-based learn- driving problem was outlined with a professional gardener
ing approach was then used to develop innovative solutions and a caretaker team. The course started by brainstorming
while validating their relevance in collaboration with gar- solutions to a problem and transitioned to identifying and
deners that were taking care of the greenhouse. As a re- outlining projects that the students chose for themselves.
sult, a number of practical applications for monitoring the Our motivation was to increase the students’ capabilities
state of the greenhouse were developed along with new prac- to work in diverse environments. Therefore, students were
tices for its maintenance. As participants were given the allowed to choose their individual projects’ hardware and
freedom to choose both the topic and technologies to work software architecture independently from a selection of very
with, the course provided a learning experience that was heterogeneous options. While each of their solutions were
tailored to suit personal interests and competences. Having different, our goal was to make them work together as a
the common background story allowed students to practice holistic system to provide a solution for the initial prob-
teamwork skills and collaborative software engineering in the lem. The ACM 2013 Computer Science Curricula empha-
context of the emerging topic of Internet of Things. sizes that the education offered to students should prepare
them for the workforce in a more holistic way than simply
conveying technical facts, and highlights the importance of
Categories and Subject Descriptors soft skills such as teamwork and problem solving as well as
K.3.2 [Computers and Education]: Computer and Infor- attributes such as sense of social responsibility [16]. We feel
mation Science Education; J.9.F [Mobile Applications]: that our learning experiment was especially well suited for
Wireless Sensor Networks; D.2.m [Miscellaneous]: Rapid this purpose.
prototyping This article is organized as follows. Section 2 outlines
the relevant technological and pedagogical concepts. Sec-
tion 3 presents the experiment setting, its position in the
Keywords curriculum, as well as student deliverables, including hard-
problem-based learning; project-based learning; internet of ware prototypes, and individual learning goals. Section 4
things; greenhouse discusses the lessons learned from the course arrangement
and reflects on the knowledge that the course staff, students,
1. INTRODUCTION and the greenhouse care team gathered during the project,
and finally, Section 5 concludes this article.
Computing has become more pervasive and embedded sys-
tems are becoming a natural part of our lives [11, 14]. While
applied computer science provides many tools for other disci-
2. BACKGROUND
2.1 Technologies
Permission to make digital or hard copies of all or part of this work for personal or
classroom use is granted without fee provided that copies are not made or distributed
Internet of Things (IoT) is an umbrella term that is used
for profit or commercial advantage and that copies bear this notice and the full cita- to describe a technological paradigm where everyday ob-
tion on the first page. Copyrights for components of this work owned by others than jects are enclosed with small, uniquely identifiable, inter-
ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or re- net enabled hardware components [1]. With these so-called
publish, to post on servers or to redistribute to lists, requires prior specific permission “Smart Devices”, product developers seek to help people in
and/or a fee. Request permissions from Permissions@acm.org.
their daily tasks, and to create new ways for working and
SIGCSE’15 March 04 - 07 2015, Kansas City, MO, USA
Copyright 2015 ACM 978-1-4503-2966-8/15/03 ...$15.00.
communicating. Smart Devices are predicted to shape the
http://dx.doi.org/10.1145/2676723.2677262. technological landscape of the near future in a radical man-
ner. A recent study by Ericsson [9] forecasted that by the The “Exact Greenhouse”-course was carried out as a seven-
year 2020 there will be 50 billion objects that are connected week project. In project-based learning students have more
to the Internet. It is expected that this movement will pro- freedom to choose the task that they will be working on and
vide new business opportunities for hardware manufacturers the domain is at least partially known beforehand [5, 21].
and ICT companies at an unforeseen scale. Authors such as Barron et al. [2] have suggested combining
IoT devices typically contain sensors that can provide in- problem- and project-based learning in a way that students
formation about the environment they are located in. Data are provided with a problem that acts as the framework of
produced by the devices is disseminated through a wire- the project and helps students form initial knowledge, which
less Internet connection to cloud-based data storage services can be further developed in a project:
that are provided by third party companies or organizations. For example, by beginning with a simulated prob-
In addition to storing the data, these services allow their lem, students develop a level of shared knowledge
users to analyze the data application programming inter- and skill that prepares them to undertake actual
faces (API) that allow creating solutions at ease, allowing projects. By following the problem with a project,
very short development cycles. students are likely to develop more flexible levels
of skills and understanding. [2]
2.2 Pedagogical Underpinnings
The experiment described in this article builds on the con- 2.3 Related Courses in Literature
structivist view of learning that suggests that people ac- Even though little work can be found on experiments that
tively construct knowledge rather than receive and store it combine computer science with horticulture, a body of aca-
(see e.g. [8, 12, 4]). Constructivists view knowledge as some- demic knowledge exist on educational experiments with a
thing that cannot be copied from what a teacher said, define cross-disciplinary approach for IoT innovation. The stud-
it as unique to those who have constructed it, and, as a con- ies underline student empowerment and aim at increasing
sequence, promote student-centered, active learning models motivation and self-direction.
of instruction. Many of these views stem from the work A number of studies emphasize the practicality of IoT
by Dewey, who wrote that “probably the greatest and com- courses (see e.g. [17, 6]). Another part of related litera-
monest mistake that we make is to forget that learning is a ture describes settings where students’ prior knowledge is
necessary incident of dealing with real life situations” [7]. low and the curriculum of the course is driven by tech-
According to Dewey, personal experience can not be substi- nological motives. Kortuem et al. [13] describe an online
tuted by even the most sophisticated learning materials. course where students were expected to buy a small hard-
Our experiment started with a problem-based set up. In ware component and to compose a simple program that uses
problem-based learning students are sent to solve realistic the Scratch [20] drag-and-drop programming environment.
and substantial problems [15, 3]. The problem creates focus Another educational experiment by Callaghan [6] introduced
for the project, stimulates the student’s thinking, position- modular Buzz-Boards for students as the sole platform for
ing the individual at the epicenter of the learning experience. creating interconnected IoT devices. Learnings from these
According to Norman, instructors should provide only lim- experiments emphasize creating holistic constructs rather
ited help in form of guidance in order to allow the students than concentrating on details of the technology [6].
to develop domain-specific knowledge and effective problem- Osipov and Riliskis describe a learning experiment that
solving skills. While students engage in the problem inde- did not involve hardware. In their example, graduate stu-
pendently, they are guided to find their own thinking strate- dents worked over a span of three courses that had a com-
gies and ways of working which improves their self-directed mon IoT theme [17]. In this example, students developed
learning skills [15]. Barrows [3] outlines five guidelines for cloud-based software applications while using a simulator to
teachers who want to use this approach in their tuition: replace physical hardware components. Student experiences
were described as mostly positive and enthusiastic, while
1. The teacher must provide a driving question that stu- some had perceived the workload as high.
dents seek answers to, without any predetermined out- When comparing the related research with our experi-
come. ment, most of the IoT-related courses are organized around
an existing set of hardware components or simulators that
2. Learning objectives and goals need to be centered around
are provided to the students, and the problems have been
the problem.
typically designed based on technologies. In our experiment
3. Students should approach the problem through con- the problem came first, and many of the components that
structive investigations. were used in the course were ordered during the course based
on the individual needs of the students, creating a novel ap-
4. They should work in an autonomous manner. proach for IOT education.
5. Projects need to be realistic so that students can make
connections to authentic situations and draw from their 3. THE GREENHOUSE PROJECT
past experiences The Exact Greenhouse is a research laboratory that is lo-
cated on the roof terrace of the Department of Computer Sci-
The problem-based learning theory is supported by the In- ence of the University of Helsinki. It was originally set up for
ductive learning theory, which is defined by Prince et al. [19] Green ICT research where the goal was to study the possibil-
as an experience where students are given a starting point ity of cooling computer servers with unconditioned outside
from which they gather observations that drive their further air, while harvesting exhaust heat into a greenhouse dur-
actions. As the insight individuals gain is a highly personal ing winter time. This setup has previously been described
interpretation, their learning pathways become diversified. in [18].
3.1 Course Setup and Curriculum 3.1.4 Week 6: Ensuring integration
Our seven week project course had a tangible and well A second milestone was set to ensure that all projects had
framed narrative that was initially presented to the students found a means to make data yielded from their IoT devices
in the course description. The criteria for student selection available into a human-readable form. The students had to
was that the students should be MSc students who were show that their prototypes were connected to the Internet
at the very end of their studies. A total of thirteen stu- and that data they produced was stored in a cloud service.
dents were selected from 25 enrolled. Limiting the amount
of participants was done to ensure meaningful learning ex- 3.1.5 Week 7: Wrap-up
periences and to facilitate studying the course structure and A course wrap-up meeting was arranged on July 14th,
organization. where the students demonstrated their project deliverables.
Two instructors were assigned to team-teach the course. Having learned about IoT technologies and seeing the so-
One of the instructors was a PhD student in Software Sys- lutions of others, the students were able to compare differ-
tems Engineering, and the other a recent PhD graduate ences of technologies and to contemplate on how choosing
with knowledge on networking and relevant IoT technolo- an another microcontroller environment could possibly have
gies. The curriculum of the experiment consisted of a kickoff changed their workflow. As their last milestone, students
meeting, three milestones and wrap-up meeting where stu- were expected to document their project outcomes and re-
dents presented their project outcomes. The course commu- flect on their learning in the course blog.
nication was arranged mainly on a course Facebook page.
The majority of the students were present in an internet 3.2 Learning Objectives and Assessment
(IRC) chat room that had originally been created for the As is normal for problem-based learning, the learning ob-
greenhouse maintenance team’s internal communication. Vol- jectives of the course were not stated at the start of the
untary weekly meetings were arranged, yet only a few stu- course. However, it was determined beforehand by the course
dents felt the need for attending them. instructors that students should learn IoT concepts, learn
how their skills can be used in places where one does not
3.1.1 Week 1: Discovering the domain typically think about it, and that the students should also
The course narrative was elaborated in full detail to the be able to design and develop a small solution that helps
students on their first meeting on Monday, June 2nd. Par- the greenhouse management at least in some fashion. No
ticipants were taken to visit the greenhouse with one of the restrictions were put on how students approached the prob-
gardeners to observe and discuss characteristics of the struc- lem, outside the rough course schedule that was outlined in
ture and needs for plant care. Afterwards, they were asked the previous section; students were naturally encouraged to
to pick a theme for their projects. Students brainstormed collaborate as well as help each other.
problems in maintaining the greenhouse and sketched solu- The course was graded pass/fail. In order to pass, stu-
tions for solving them. Numerous project ideas emerged, dents had to pass three milestones (weeks 1, 2 and 7), demon-
including topics such as “Smart Flowerpot”, “Hydroponics”, strate their solution that consisted of a working hardware
“Temperature” and “Ventilation” among others. prototype and source code, and provide the sources to their
Next, participants were introduced to their solution do- project via a version control system (in our case, GitHub1 ).
main: a set of microcontroller equipment that had been re- The assessment approach was chosen due to the variety of
served for the course. They were encouraged to choose any different paths that students could take as well as due to the
platform, including but not limited to the devices that were open-ended nature of the project.
available at the moment. After the meeting participants
were set off with the hardware and their first task: to get to 3.3 Resources and Budget
know the technology and to identify components that were While no fixed cost budget was set up for the course, re-
required for their project implementation. sources were very limited and thus inexpensive microcon-
troller platforms such as the Arduino2 , Raspberry Pi3 , Elec-
3.1.2 Week 2: Creating project requirements tric IMP4 , and Intel Galileo5 were preferred over out-of-
the box, programmable smart greenhouse devices that were
On their second meeting on Monday 9th, students met a available on the market. However this did not limit the stu-
professional gardener who further outlined problems related dents. Even though they could design their hardware archi-
to greenhouse maintenance and provided feedback on the rel- tecture quite freely, the total cost of the course was approx-
evance of the students’ plans. The feasibility of some of the imately 500 US dollars. For application development and
project ideas were also discussed with the course staff and, data storage, students used cloud-based service providers
finally, a compromised list of components to order was com- that offered free access for small-scale developers6,7 . Ta-
posed. Orders were placed immediately from manufacturers ble 1 summarizes the types and cost of components that
that were known for fast delivery. The second assignment were used in individual projects.
was to document their full plans to a course blog. This was
the first milestone of the course. 1
http://github.com
2
http://www.arduino.cc
3.1.3 Weeks 3-5: Autonomous project work 3
http://www.raspberrypi.org
4
During the following weeks, students built their hard- http://www.electricimp.com
5
ware prototypes and developed related software. Voluntary http://arduino.cc/en/ArduinoCertified/
weekly meetings were arranged on Mondays and the course IntelGalileo
6
staff provided technical help on demand. http://www.heroku.com
7
http://www.firebase.com
 Table 1: The overall cost of the projects.

Project 1: Visualization
Arduino Duemilanove $20
DHT-11 temperature humidity Sensor $5
$25

Project 2: Humidity A
Arduino Uno $25
SeedStudio Grove - Moisture Sensor $5
Simple photoresistor $0
Actuators: buzzer, RGB LED $5
$35

Project 3: Humidity B
Electric Imp and breakout board $45
SeedStudio Grove - Moisture Sensor $5
SeedStudio Soil Moisture and Temp. Sensor $40 Figure 1: High level architecture of the Greenhouse
$90 Care project.
Project 4: Temperature IoT device into a cloud-based storage and provided a chart
Intel Galileo $100 representation. Other projects utilized this application for
DS18B20 temperature sensors + cables $55 either visualizing their own data or yielding data for their
$155 own purposes. The few students that did not, created their
own user interfaces for displaying their data according to
Project 5: Air Flow their own plans.
Arduino Uno $25
Computer case fans $5 3.4.2 Humidity (2,3)
Power source 12v $5
The most recurring greenhouse care task is to check for
Prototyping breadboards $10
humidity of soil and add water and fertilizer if needed. Two
Misc. electronics(resistors, transistors) $5
student projects (2,3) approached facilitating this task by
$50
creating independent solutions for sensing humidity inside a
planting container. The main emphasis of the first project
Project 6: Hydroponics
was on interacting with people and creating alarms to in-
Arduino YÚN $75 dicate when it was time to add water the soil. The other
Stackable PCB headers $2 product focused on product innovation: how a “smart flow-
Seedstudio One Wire Temperature Sensors $12 erpot” could be transformed into a commercial product.
Seedstudio Grove - UV Sensor $10
Seedstudio Grove Base Shield $18 3.4.3 Temperature and air (4,5)
Grove humidity and temperatur pro sensors $15
At times a space heater was needed to heat the green-
Grove air quality sensor $10
house. While the greenhouse was custom-built from inex-
$143
pensive components, its ventilation was compromised and
heat distributed unevenly. The “Temperature” project (4)
Project 7: Visitors
concentrated on monitoring six different parts of the green-
Raspberry Pi with WiFi dongle $50
house with air temperature sensors and the “Air Flow” project
Raspberry Pi Board For NFC-Tags $40
(5) utilized this data to help circulate heat more evenly
MiFare Classic NFC Keychain Fobs $13
within the greenhouse.
$103
3.4.4 Hydroponics (6)
Total $510 One of the installments in the greenhouse was a tank
where plants were grown in nutrient solution, without us-
3.4 Student projects ing soil. The system was not working correctly, leading to
In this chapter, we give an overview of the student projects. plants dying after a few weeks of their installation. One
The examples are expected to provide an authentic and prac- student aimed at building a measurement device for this
tical representation on both the scale and scope of the ex- small-scale ecosystem. The solution consisted of an imagi-
periment. native amount of sensors that could be used for measuring
both air and water quality.
3.4.1 Visualization (1)
To be able to support the work of gardeners, all projects 3.4.5 Visitors (7)
were required to have a human readable user interface. This The question “Has anybody visited the greenhouse?” was
was first achieved by project 1 in which the student com- one of the most common phrases in the communication of
posed a web application that allowed pushing data from any the caretaker team. Therefore, a status reporting device was
created, allowing visitors to the greenhouse to flash Near 4.2.1 Project 1: Visualization
Communication Field (NFC) enabled keychains to a reader Both IoT technologies and web applications were a part
located inside the greenhouse. A requirement for the system of this student’s professional career. He was very active and
was that it should be usable with wet and dirty hands. took a central role in the project and offered his work for
others to use, too.
4. DISCUSSION 4.2.2 Project 2: Humidity A
After the initial problem-based learning phase, students This student had done some microcontroller-based projects
were given the freedom to choose their own project topic and before the course. In this project, he struggled with hard-
scope. In this section we first reflect on how this freedom ware organization and thus learned about physical design.
affected solution architecture design and consequently, the The merit of his work was that the result was strongly con-
course budget. Finally, we contemplate the expectations and nected to human interaction and solving a concrete problem.
learning outcomes of both students and course personnel.
4.2.3 Project 3: Humidity B
4.1 Budget Guidance This team consisted of two students who chose a stream-
During the first meeting the instructors suggested micro- lined technology, making their solution almost effortless to
controller platforms that could be chosen as a starting point compose. The merit of their result was in their ability to
for projects. We did not communicate how much each indi- see their result as a commercial product for the near future
vidual project should cost and emphasized that the options development of which they had a clear plan. They never
were not limited to the suggested technologies. Participants seemed to leave their comfort zone to learn about the IoT
became inspired by the possibilities. Some envisioned con- technology itself.
cepts that were impossible to implement within the time-
frame and some requested high-cost, professional sensors. 4.2.4 Project 4: Temperatures
The people who were guided to cut back their plans, natu-
This student had excellent background knowledge in IoT
rally, expressed their disappointment. However, the negative
technologies. His output was not a learning experience, but
emotions subsided after they received the more affordable
proof of learning that had happened before the course. He
hardware components we had helped them to choose. We
worked very independently, without much communication
believe that if a finite cost limit would have been given for
with others.
each student project, it might have helped to mitigate the
initial plans. However, having a fixed budget might have led
4.2.5 Project 5: Air Flow
to less inspired plans and at worst, students trying to fill up
their budgets with components that they might not have a The student was very eager to learn new things and pro-
need in the end. gressed through trial and error. Even though many of his
plans were never realized, he retained his enthusiasm and
4.2 Grading the Individual Learning Process presented a wide array of new knowledge.
After students had completed their demos, the instructors 4.2.6 Project 6: Hydroponics
reviewed the projects and gave students a simple pass or
fail grade. In this section, we present the highlights of each The student eagerly pulled together a meritious sensor
student’s learning pathways. Table 2 summarizes learning system with many hardware components. However, it re-
themes the instructors identified each student had touched mained unclear whether he had understood the purpose of
with while their project work. his system, and its relation to its users.

4.2.7 Project 7: Visitors

Table 2: The learning theme matrix. This student had no prior experience with microcontrollers.
P1 P2 P3 P4 P5 P6 P7 He was very active in the design stage of his system and
Hardware showed understanding of the requirements engineering pro-
Microcontroller technology • • • cess. When it was time for implementation, he isolated him-
Sensor technology • • • self and was absent a lot. The resulting system was designed
Hardware integration • • to fulfill a clear purpose and was technologically sound.
Physical design (3d) •
Testing • 4.3 Unexpected learning outcomes
For the whole duration of the course, students were partic-
Software ipating in the greenhouse caretaker team’s online chat. The
Web application development • • • encounters encounters led to discussions on what type of
Microcontroller programming • • • • • • knowledge would be useful for facilitating the team’s work.
Software integration • • • • • They also created new knowledge for the maintenance team
on how their inter-person collaboration could be facilitated.
Other Contrary to our expectations, we quickly observed that
Product development • prototyping simple ideas using microcontrollers and connect-
Group work • • • • ing them to the internet was easy for the participants. As
devices were built and left running for a time, we discovered
that the microcontrollers used plenty of electricity consider-
ing their size and functionality. Even the simplest prototype
 measuring the level of lighting of a room emptied its batter-
ies in a few days. This led us to discussions on energy effi- [3] H. S. Barrows. Problem-based learning in medicine
ciency and the carbon footprint of the devices that we use and beyond: A brief overview. New Directions for
in everyday life. The students explored various optimiza- Teaching and Learning, 1996(68):3–12, 1996.
tion strategies such as disconnecting the components from [4] M. Ben-Ari. Constructivism in computer science
the internet for longer periods and activating components education. Journal of Computers in Mathematics and
only when needed. While doing so, they also learned to Science Teaching, 20(1):45–73, 2001.
consider issues such as measurement accuracy. These topics [5] P. C. Blumenfeld, E. Soloway, R. W. Marx, J. S.
were new to both students and the course staff. Krajcik, M. Guzdial, and A. Palincsar. Motivating
At the wrap-up meeting, the students were expected to project-based learning: Sustaining the doing,
present their projects. As some students were absent, they supporting the learning. Educational psychologist,
were allowed to present their work to a course instructor 26(3-4):369–398, 1991.
separately. Those that were not able to attend the get- [6] V. Callaghan. Buzz-Boarding; practical support for
together, missed a great peer-to-peer discussion while par- teaching computing based on the internet-of-things.
ticipants compared their solutions that shared a similar pur- 2012.
pose, but were built on very different technologies. As one [7] J. Dewey. The School and the Society: Rev. Ed.
of the students put it: University of Chicago Press, 1967.
During the demo session, I understood how dif- [8] R. Driver and B. Bell. Students’ thinking and the
ferent the solutions were both from the hardware- learning of science: A constructivist view. School
and software-perspective. I would not have been science review, 67(240):443–56, 1986.
able to compare technologies without my own project, [9] Ericsson. More than 50 Billion Connected Devices -
as my knowledge was not at a level that I could Taking Connected Devices to Mass Market and
have talked about it in a meaningful way. Profitability. White paper 284 23-3149 Uen, 2011.
[10] K. A. Frenkel. Computer science meets environmental
science. Commun. ACM, 52(9):23–23, Sept. 2009.
5. CONCLUSIONS [11] A. Galloway. Intimations of everyday life: Ubiquitous
In this article, we have described an experimental low-cost computing and the city. Cultural Studies,
course where students develop Internet of Things applica- 18(2-3):384–408, 2004.
tions for improving the maintenance of a greenhouse. The [12] T. Greening. Emerging constructivist forces in
course was organized in two phases. First, a problem-based computer science education: Shaping a new future? In
learning approach was used to learn about the environment, Computer science education in the 21st century, pages
identify problems and to come up with solutions for im- 47–80. Springer, 2000.
proving them. In the second phase, the students worked on [13] G. Kortuem, A. K. Bandara, N. Smith, M. Richards,
projects, where these solution ideas were made real. The and M. Petre. Educating the internet-of-things
outcome of this project contrasted the wide array of pos- generation. Computer, 46(2):53–61, 2013.
sibilities that blending problem- and project-based learning [14] M. Kranz, P. Holleis, and A. Schmidt. Embedded
approaches can provide for students and teachers. interaction: Interacting with the internet of things.
None of the course instructors and students had previous Internet Computing, IEEE, 14(2):46–53, 2010.
experience of maintaining a greenhouse. Overall, the expe-
[15] G. R. Norman and H. G. Schmidt. The psychological
rience did not only teach students about technical aspects
basis of problem-based learning: a review of the
that were related to the Internet of Things, but also pro-
evidence. Academic medicine, 67(9):557–65, 1992.
vided viewpoints on urban sustainability and Green ICT.
[16] A.-C. J. T. F. on Computing Curricula. Computer
The project also demonstrated to the department that the
science curricula 2013. Technical report, ACM Press
organization of such loosely defined projects with little pre-
and IEEE Computer Society Press, December 2013.
defined hardware and a relatively small budget can provide
both a successful and valuable learning experience. Both the [17] E. Osipov and L. Riliskis. Educating innovators of
gardeners and the students in the course gained experiences future internet of things. In Frontiers in Education
that are likely to benefit their future, many of them praising Conference, 2013 IEEE, pages 1352–1358. IEEE, 2013.
the course as an eye-opening experience. Due to the success [18] M. Pervilä, L. Remes, and J. Kangasharju. Harvesting
and demand of the course, we are planning to repeat the heat in an urban greenhouse. In ACM SIGMETRICS
course in the future, and will especially focus on approaches Performance Evaluation Review, volume 41, pages
for introducing the course to students who are at an earlier 95–97, 2013.
stage in their studies. [19] M. Prince and R. Felder. Inductive teaching and
learning methods: Definitions, comparisons, and
research bases. Journal of engineering education,
6. REFERENCES 95(2):123–138, 2006.
[1] L. Atzori, A. Iera, and G. Morabito. The internet of [20] M. Resnick, J. Maloney, A. Monroy-Hernández,
things: A survey. Computer networks, N. Rusk, E. Eastmond, K. Brennan, A. Millner,
54(15):2787–2805, 2010. E. Rosenbaum, J. Silver, B. Silverman, et al. Scratch:
[2] B. J. Barron, D. L. Schwartz, N. J. Vye, A. Moore, programming for all. Communications of the ACM,
A. Petrosino, L. Zech, and J. D. Bransford. Doing 52(11):60–67, 2009.
with understanding: Lessons from research on [21] J. W. Thomas. A review of research on project-based
problem-and project-based learning. Journal of the learning. 2000.
Learning Sciences, 7(3-4):271–311, 1998.

View publication stats