Content uploaded by Theodoros Thodoris Anagnostopoulos
Author content
All content in this area was uploaded by Theodoros Thodoris Anagnostopoulos on Jun 27, 2015
Content may be subject to copyright.
adfa, p. 1, 2015.
© Springer-Verlag Berlin Heidelberg 2015
Waste management as an IoT enabled service in Smart Cities
Alexey Medvedev1, Petr Fedchenkov1, Arkady Zaslavsky2,
1
,
Theodoros Anagnostopoulos3,1, Sergey Khoruzhnikov1
1ITMO University, Kronverkskiy pr., 49, St.-Petersburg, Russia
alexey.medvedev@niuitmo.ru
petr_fedchenkov@niuitmo.ru
xse@vuztc.ru
2CSIRO, Melbourne, Australia
arkady.zaslavsky@csiro.au
3Community Imaging Group, University of Oulu, 90570 Oulu, Finland
tanagnos@ee.oulu.fi
Abstract. Intelligent Transportation Systems (ITS) enable new services within Smart
Cities. Efficient Waste Collection (WC) is considered a fundamental service for Smart
Cities. Internet of Things (IoT) can be applied both in ITS and Smart cities forming an
advanced platform for novel applications. Surveillance systems can be used as an as-
sistive technology for high Quality of Service (QoS) in waste collection. Specifically,
IoT components: (i) RFIDs, (ii) sensors, (iii) cameras, and (iv) actuators are incorpo-
rated into ITS and surveillance systems for efficient waste collection. In this paper we
propose an advanced Decision Support System (DSS) for efficient waste collection in
Smart Cities. The system incorporates a model for data sharing between truck drivers
on real time in order to perform waste collection and dynamic route optimization. The
system handles the case of ineffective waste collection in inaccessible areas within the
Smart City. Surveillance cameras are incorporated for capturing the problematic areas
and provide evidence to the authorities. The waste collection system aims to provide
high quality of service to the citizens of a Smart City.
Keywords: Waste Collection, Smart City, Internet of Things (IoT), Intelligent Trans-
portation Systems, Surveillance systems
1 Introduction
Recent advances in production of mobile computers and smartphones, smart sensors and
sensor networks in connection with next generation mobile networks opened vast opportuni-
ties for researchers and developers of various systems and application in the field of Smart
Cities and Intelligent Transportation Systems (ITS). Thought some areas like application for
1
Arkady Zaslavsky is an International Adjunct Professor at ITMO University since 2012
monitoring public transport are already well researched, other areas are still working with
outdated technologies and models. One of such areas is the management of solid waste col-
lection process. In a Smart City collection of waste is a crucial point for environment and its
quality should be considered seriously. In order to understand the concept of Smart Cities in
depth, a suitable definition is provided. In this research we use the most suitable definition
for the IoT-enabled waste collection in Smart Cities, which is [1]: “A Smart City is a city
well performing in a forward-looking way in the following fundamental components (i.e.,
Smart Economy, Smart Mobility, Smart Environment, Smart People, Smart Living, and
Smart Governance), built on the ‘smart’ combination of endowments and activities of self-
decisive, independent and aware citizens”. In this definition we can see important compo-
nent - Smart Environment - which is tightly connected to environmental pollution. The main
countermeasure to environmental pollution in terms of a Smart City is the IoT-enabled
waste collection. The following definition of IoT is used in this paper [2]: “The Internet of
Things allows people and things to be connected Anytime, Anyplace, with Anything and
Anyone, ideally using Any path/network and Any service”. IoT technologies enable new
services and reshape the existing ones in Smart Cities [3]. For instance static waste collec-
tion is redesigned to Waste Collection as a Service. As the result this enables online dynam-
ic scheduling and routing of the trucks [4]. Issues connected to dynamic waste collection
could be divided into 2 main problems: (i) when to collect waste form bins (i.e., schedul-
ing), and (ii) what route the trucks will follow (i.e., routing).
In this paper we propose waste collection system enhanced with IoT services which enable
dynamic scheduling and routing in a Smart City. We also present the design of a cloud sys-
tem for organization of waste collection process and applications for waste truck drivers and
managers. The proposed system also features an on-board surveillance system which raises
the process of problem reporting and evidence collection to a higher level.
The rest of the paper is structured as follows. Section II presents related work in the area of
IoT-enabled waste collection in Smart Cities. Section III describes the main features of the
system and some scenarios of usage. Section IV presents the system model architecture and
two applications. One is a mobile application for the waste truck driver and another is a web
application for waste management company. Section V contains the evaluation on one pos-
sible scenario and section VI concludes the paper and discusses future work.
2 Related work
The area of route planning and optimizing for logistic purposes is well-researched and hun-
dreds of Intelligent Transportation Systems already exist. There are also a number of pro-
jects aiming to provide an effective system specializing on waste collection needs. A Geo-
graphical Information System (GIS) transportation model for solid waste collection that
elaborates plans for waste storage, collection and disposal has been proposed in [5] for the
city of Asansol in India. In [6] an enhanced routing and scheduling waste collection model
is proposed for the Eastern Finland, featuring the usage of a guided variable neighbourhood
thresholding metaheuristic. In the city of Porto Alegre in Brazil authors propose [7] a truck-
scheduling model for solid waste collection. The aim of the research was to develop an op-
timal schedule for trucks on defined collection routes. Examples of other systems are de-
scribed in [8],[9],[10] and [11].
A survey presented in [12] reviews the researches done on waste collection in developing
countries from 2005 to 2011 and considers challenges for developing countries in waste
collection sphere. The research focuses on determination the stakeholders’ ac-
tions/behaviour and evaluation of influential factors defining their role in waste collection
process. The models in the survey were tested on real data. In [13] a survey considering
system approaches for solid waste collection in developing countries is presented. The re-
search compares the history and the current practices, presented from 1960s to 2013. Infor-
mation about the challenges and complexities is also given there. The output of the survey is
drawing a conclusion that developing and implementing solid waste collection approaches
in developing countries are of a great importance. The main issue is that waste collection
does not include innovation that IoT can provide. Models do not use real time information
of the waste collection, although some approaches use advanced scheduling and routing via
exploiting modern ICT algorithms. Information about bins status was not considered as part
of waste collection. All the reviewed surveys do not propose a model that will use IoT tech-
nology for Smart Cities, though they consider different approaches for waste collection.
Moreover, enabling combined participation of stakeholders like road police and city admin-
istrations in one system is not covered. Finally, the concept of implementing an on-board
surveillance system for fast problem reporting and evidence collecting is not implemented in
mentioned projects, but is described in [14] and [15] separately from waste management
topic. All this allows stating the need for development a system facilitating the usage of IoT
data, dynamic routing models and participation of diverse types of stakeholders.
3 Main features and scenarios of usage
System architecture aims to suit two main targets. First target is providing software-as-a-
service (SaaS) products for customers. Mainly, these customers are private companies that
are involved in waste collection, owning waste trucks, organize work of drivers, get con-
tracts from municipalities and pass wastes to recycling organizations or city dumps. Second
main target is developing a system, which makes possible mutually beneficial communica-
tion between all the stakeholders involved in the chain of supplying goods and utilizing
solid waste in smart city.
Fig. 1. The big picture of a waste collection management system
A list of possible stakeholders of the system and brief description of their needs, business
rules, possibilities and connections with others is presented below:
City administration needs understanding of the big picture, generating reports, control
over pricing etc.
District administrations are interested in controlling the process of waste collection,
checking quality of service (all waste collected, all in time, waste collected cleanly, waste
transported to special places), quick and legal ways for solving disputes and problems.
Municipalities can also deploy and maintain smart city infrastructure like capacity sen-
sors in waste bins and wireless networks for data transferring.
Waste trucks owning companies need platform for organizing and optimization of their
business process in general without serious investments in developing, deploying and
supporting their own system. Such system must include effective dynamic routing based
on IoT data for the truck fleet. Besides, controlling drivers and tracking the fleet is also
an important issue.
Waste truck drivers need navigation system for fulfilling their tasks. Another issue is
reporting problems and passing them to the operators in the office instead of thinking
how to solve the problem, this can sufficiently save time of a driver and vehicle. Drivers
also need evidence that their work was done correctly and cleanly.
Managers of dumps and recycling factories can publish their possibilities or needs in
acquiring certain amount of waste for storing or recycling
Staff that is responsible for trash bins in the current yards needs communications with
waste management companies and truck drivers.
Road police can get reports about inaccurate car parking that leads to impossibility of
waste collection.
Citizens want to have better service, lower cost and having easy accessible reports on
what has been done and how much it cost.
As it is shown on fig. 1, the main component of the system is the cloud based decision sup-
port system (DSS). It is a platform that provides possibilities for intercommunication of all
the stakeholders. Waste trucks generate sensor data about their capacity, location, fuel avail-
able and consumed. Besides, truck drivers load video fragments or pictures of problems
they meet while performing their tasks. Sensors located in smart bins generate data about
capacity, pollution etc. Waste management companies after registering in the system create
rules and business logic for waste collection. Creating the business logic and rules means
registering the companies’ fleet and drivers, registration of smart and non-smart bins from
which waste must be collected, defining time windows for waste collection that corresponds
to local laws and terms of contract with the municipality. Important issue is gathering, pro-
cessing and storing data from heterogeneous sensors, including capacity sensors in bins and
trucks, cameras, Internet connected objects (ICOs), etc. On-line navigation systems provide
data about traffic situation, which is crucial for effective routing. It is much more convenient
and cost effective to use this data from special services using sensing-as-a-service [16] mod-
el, rather than implementing such function inside the DSS system.
Having all this information in DSS it becomes possible to provide customers with best pos-
sible routing for each truck. Moreover, reports from drivers when they encounter a problem
on the road are processed semi-automatically that results in a much faster problem solving.
It is possible to count plenty of system usage scenarios but due to the lack of space we pre-
sent and evaluate only one, which is presented in the “Scenarios” section of this chapter.
3.1 Scenario – inaccessible waste bin
Waste truck drivers report about his inability to drive inside the yard or approach the waste
bin with the truck and get wastes. Usual reason for it is inaccurate car parking, which is
shown in fig. 2.
Fig. 2. – Inaccessible waste bin scenario
This report includes video or picture of a problem made with drivers’ smartphone or tablet
on which the client android app is running. This data is annotated by a voice message, GPS
coordinates and other available data. Then this report is processed in the DSS and if it is
correct it is sent to organizers of waste collection in this particular place and to the road
police. The truck driver doesn't waste time for waiting, he/she goes to the next point and the
route is dynamically recounted. When the problem is solved the system recounts the route
for one of the available trucks and the waste from unlocked bin is collected. It is combined
with dynamic routing algorithms [17] to maximize the efficiency of waste collection. As it is
stated in [17] static models do not fit the idea and IoT-enabled potentiality of a smart city. It
is often faster and cheaper to make a longer route saving time from traffic congestions or
waiting for a road problem to be solved; thereafter the need for IoT-enabled dynamic rout-
ing engine for the fleet becomes one of the main features of the designed DSS. Schematical-
ly old static and new dynamic approaches are shown on fig. 3.
Fig. 3. – Static and dynamic routing approaches
4 System Model Architecture and Applications
As most Intelligent Transportation Systems, the designed system also implements the engine
for storing, rendering, updating and displaying maps as one of the main components. Some
of the criteria for choosing the engine were licence independence, possibility for making
changes in existing maps and possibility to build a separate instance in a private own cloud.
As a result OpenStreetMap [18] has been chosen as the main technology for acquiring maps
data and for displaying maps and routes both for the drivers’ android application and web
application for managers and other clients. Nominatim [19] is a part of OpenStreetMap pro-
ject; it is used for geocoding – finding latitude and longitude by in OSM data by name and
address.
As it was already mentioned above, a typical client of the described system is a waste man-
agement company that owns a heterogeneous fleet of vehicles and needs to service a number
of points in a city. This is a well-known problem in logistics and transportation - the vehicle
routing problem (VRP) [20] and its objective is to minimise the total route cost. There are
several variations and specializations of the VRP but their description is omitted in this pa-
per due to space limitations. A number of open source and commercial projects exist ena-
bling fast solution of VRP. Examples of such projects are JSPRIT [21], Open-VRP, Opta-
Planner, SYMPHONY, VRP Spreadsheet Solver etc. JSPRIT [21]–java based, open source
toolkit for solving rich traveling salesman (TSP)[26] and vehicle routing problems (VRP)
has been chosen the main library used for solving VRP and building initial routes due its
lightness, flexibility and ease of use. Another advantage of JSPRIT library is its easy exten-
sibility that will be significantly useful while adding special features and algorithms specific
for waste collection. GraphHopper [27] is a fast and memory efficient Java road routing
engine. It is used for calculating optimized routes for waste trucks based on OpenStreetMap
data.
A web-based application for waste management companies is presented in fig. 4. It provides
managers and operators with facilities like registering the infrastructure and vehicles, track-
ing the fleet, mark waste bins as blocked and unblocked etc. A mobile Android-based appli-
cation for a waste truck driver is shown in fig. 7. As the main feature it delivers smart navi-
gation options to the driver. Secondly, the application provides an option of reporting a
problem. In fig 7. (right) a process of making a report about a blocked by car waste bin is
shown. Noticeably, we implement a feature of annotating a report with voice that allows not
distracting the driver from his work.
Fig. 4. – Web-based application for waste management companies
4.1 Surveillance System
As it was already mentioned in Introduction section, one of the main features of the pro-
posed system is a waste truck based surveillance system that serves several purposes. These
purposes are:
Evidence collection system for easy accident analysis [14]
Reporting road and other problems [15]
Proof of correctly and cleanly done work
First two scenarios are based on CityWatcher application and are described in [14] and [15].
CityWatcher is an android based application for smartphones, which acts as an IoT car black
box. It records video of the situation on the road and annotates it with time and coordinates.
The difference with other black boxes is the ability of authorized personnel to make requests
to local storage of all participating in the system devices to search for an evidence of road
accidents.
The distinction from CityWatcher application is in the number of cameras used. CityWatch-
er was designed to use with the camera built into smartphone. This was a suitable solution
for a civil purpose, but it may be not enough for professional service. In case of waste truck
surveillance system several wired or wireless cameras can be used simultaneously.
Fig. 5. – XML file with vehicle and task description
5 Experimental Evaluation
We use real and synthetic data to evaluate the proposed system. Real data is the road graph
of St.-Petersburg and waste bins location. In the first experiment we use 6 trucks for collect-
ing waste from 24 bins. The task for the JSPRIT library algorithms is described as an XML
file, a part of which is presented on fig 5. The XML file contains the description of one ve-
hicle and one pickup point, which represents a smart waste bin.
The result of the VRP solvation is graphically presented in fig 6. As a result after the first
experiment we have routes for several trucks, distances, time and fuel consumed. This is the
best-case scenario, as all the bins in this experiment are treated as accessible (not blocked).
Fig. 6. - The result of the VRP solvation
Routes for this experiment are presented in fig. 4 (all trucks for manages) and fig 7 (one
route for a truck driver).
Fig. 7. – Navigation (left) and problem reporting (right)
In the second experiment several random bins are blocked. We use this approach several
times on different percentage of blocked bins.
When the truck driver finds a blocked bin or other problem that makes it impossible to col-
lect the waste he/she loses several minutes for reporting the problem via telephone and
leaves the place. When all accessible points are collected the driver makes one more round
for collecting waste from bins that are assumed to be unblocked now. This is the worst case
scenario, as the diver loses time for driving into the yard, recognizing the problem, reporting
it and returning to the same place later.
In the third experiment we assume that when a bin is blocked the truck reports the problem
with a mobile application and continues the trip. For example, the waste collection point in
the yard contains four bins – for plastic, glass, paper and organics. While the truck that re-
ports the problem (e.g. collecting plastic) does not get significant resource economy, other
three trucks are informed about the problem and automatically exclude current point from
their route. When the problem is marked as solved by police or municipality staff the route
is dynamically rebuilt and one of the available trucks gets a task for collecting waste from
that point. Line graph for this experiment are presented on fig 8. The lowest line represents
the ideal scenario without blocked bins. It is easy to see, that total time (and accordingly
cost) used by informed waste trucks (red line) in comparison with the scenario without in-
forming drivers about blocked bins (green line) is significantly lower. This experimental
evaluation showed that our approach for coping with blocked bins scenario is cost-effective.
Fig. 8. Dependence of total time spent by fleet for the percentage of blocked waste bins
6 Conclusion and future work
In this paper we have presented a novel cloud-based system for waste collection in smart
cities. The system aims to provide services for different kind of stakeholders involved in this
area - from city administrations to citizens. Still, the design focuses mostly on providing
SaaS services to commercial waste management companies. Development of applications
for city administrations, municipal staff, recycling factories and other stakeholders is
planned to be done in future. We have evaluated the proposed system and shown that im-
plementing on-board surveillance cameras for problem reporting in conjunction with a cloud
DSS system and dynamic routing models can give a significant increase of cost-
effectiveness, which is one of the most indicating criteria for the Smart City.
ACKNOWLEDGEMENT
The research has been carried out with the financial support of the Ministry of Education
and Science of the Russian Federation under grant agreement #14.575.21.0058.
References
1. Centre of Regional Science. Vienna University of Technology. Smart Cities. Ranking of Europe-
an Medium-Sized Cities. http://www.smart-cities.eu, Accessed on: December 23, 2014.
2. Guillemin P., Friess P., “Internet of things strategic research roadmap,” The Cluster of European
Research Projects, Tech. Rep., September, 2009, http://www.internet-of-things-research.eu, Ac-
cessed on: December 23, 2014.
3. Delicato F. C., Pires P. F., Batista T., Cavalcante E., Costa B., Barros T., “Towards an IoT eco-
system”, In the Proceedings of the 1st ACM International Workshop on Software Engineering for
Systems-of-Systems, SESoS’13, pp. 25-28, Montpellier, France, July, 2013.
4. Lingling H., Haifeng L., Xu X., Jian L., “An Intelligent Vehicle Monitoring System, Based on
Internet of Things”, IEEE 7th International Conference on Computational Intelligence and Securi-
ty (CIS), pp. 231-233, Hainan, December, 2011.
5. Ghose M. K., Dikshit A. K., S. K. Sharma, “A GIS based transportation model for solid waste
disposal – A case study on Asansol municipality”, Journal of Waste Management, vol. 26 (11),
pp. 1287-1293, January, 2006.
6. Nuortio T., Kytojoki J., Niska H., Braysy O., “Improved route planning and scheduling of waste
collection and transport”, Journal of Expert Systems with Applications, vol. 30 (2), pp. 223-232,
February, 2006.
7. Li J. Q., Borenstein D., Mirchandani P. B., “Truck scheduling for solid waste collection in the
City of Porto Alegre, Brazil”, Journal of Omega, vol. 36 (6), pp. 1133-1149, December, 2008.
8. Zamorano M., Molero E., Grindlay A., Rondriquez M. L., Hurtado A., Calvo F. J., “A planning
scenario for the application of geographical information systems in municipal waste collection: A
case of Churriana de la Vega (Granada, Spain)”, Journal of Resources, Conservation and Recy-
cling, vol. 54 (2), pp. 123-133, December, 2009.
9. Tavares G., Zsigraiova Z., Semiao V., Carvalho M. G., “Optimisation of MSW collection routes
for minimum fuel consumption using 3D GIS modeling”, Journal of Waste Management, vol. 29
(3), pp. 1176-1185, March, 2009.
10. Benjamin A. M., Beasley J. E., “Metaheuristics for the waste collection vehicle routing problem
with time windows, driver rest period and multiple disposal facilities”, Journal of Computers &
Operations Research, vol. 37 (12), pp. 2270-2280, December, 2010.
11. Son L. H., “Optimizing Municipal Solid Waste collection using Chaotic Particle Swarm Optimi-
zation in GIS based environments: A case study at Danang city, Vietnam”, Journal of Expert Sys-
tems with Applications, vol. 41 (18), pp. 8062-8074, December, 2014.
12. Guerrero L. A., Maas G., Hogland W., “Solid waste management challenges for cities in develop-
ing countries”, Journal of Waste Management, vol. 33 (1), pp. 220-232, January, 2013.
13. Marshall R. E., Farahbakhsh K., “Systems approaches to integrated solid waste management in
developing countries”, Journal of Waste Management, vol. 33 (4), pp. 988-1003, April, 2013.
14. Medvedev A., Zaslavsky A., Grudinin V., Khoruzhnikov S., Citywatcher: Annotating and Search-
ing Video Data Streams for Smart Cities Applications, Internet of Things, Smart Spaces, and Next
Generation Networks and Systems 14th International Conference, NEW2AN 2014 and 7th Con-
ferenceruSMART, pp. 144-155, 2014
15. Medvedev A., Zaslavsky A., Khoruzhnikov S., Grudinin V., Reporting Road Problems in Smart
Cities Using OpenIoT Framework, Interoperability and Open-Source Solutions for the Internet of
Things, Lecture Notes in Computer Science Volume 9001,pp 169-182, 2015
16. Perera C., Zaslavsky A., Christen P., Georgakopoulos D. Sensing as a service model for smart cit-
ies supported by Internet of Things. Transactions on Emerging Telecommunications Technologies
25 (1), 81-93
17. Anagnostopoulos T., Zaslavsky A., “Effective Waste Collection with Shortest Path Semi-Static
and Dynamic Routing”, Proceedings of the 7th International Conference ruSMART, Internet of
Things, Smart Spaces and Next Generation Networks and Systems, Lecture Notes in Computer
Science, Springer, 2014, vol. 8638, pp 95-105.
18. OpenStreetMap Webpage, https://www.openstreetmap.org, [Accessed on 03 March 2015]
19. Nominatim Wiki Webpage, http://wiki.openstreetmap.org/wiki/Nominatim [Accessed on 03
March 2015]
20. Dantzig, George Bernard; Ramser, John Hubert (October 1959). The Truck Dispathing problem,
Management Science 6 (1): 80–91.
21. JSPRIT Webpage, http://jsprit.github.io/, [Accessed on 15 April 2015]
22. Open-VRP Webpage, https://github.com/mck-/Open-VRP/, [Accessed on 15 April 2015]
23. OptaPlanner Webpage, http://www.optaplanner.org/, [Accessed on 15 April 2015]
24. SYMPHONY Webpage, https://projects.coin-or.org/SYMPHONY/, [Accessed on 15 April 2015]
25. VRP Spreadsheet Solver Webpage, http://verolog.deis.unibo.it/vrp-spreadsheet-solver/, [Accessed
on 27 April 2015]
26. TSP Webpage, http://en.wikipedia.org/wiki/Travelling_salesman_problem/, [Accessed on 27
April 2015]
27. GraphHopper Webpage, https://graphhopper.com/ , [Accessed on 27 April 2015]