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

net/publication/335241161

Blockchain Based Car-Sharing Platform

Conference Paper · September 2019

DOI: 10.1109/ELMAR.2019.8918650

CITATIONS READS

8 5,897

4 authors:

Viktor Valaštín Kristian Kostal

Slovak University of Technology in Bratislava Slovak University of Technology in Bratislava
1 PUBLICATION   8 CITATIONS    10 PUBLICATIONS   112 CITATIONS   

SEE PROFILE SEE PROFILE

Rastislav Bencel Ivan Kotuliak

Slovak University of Technology in Bratislava Slovak University of Technology in Bratislava
11 PUBLICATIONS   37 CITATIONS    83 PUBLICATIONS   359 CITATIONS   

SEE PROFILE SEE PROFILE

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

LdV project Train2Cert View project

Smart Electricity Metering System Using Blockchain and IoT View project

All content following this page was uploaded by Kristian Kostal on 19 August 2019.

The user has requested enhancement of the downloaded file.

 Blockchain Based Car-Sharing Platform
Viktor Valaštín, Kristián Košt’ál, Rastislav Bencel, Ivan Kotuliak
Faculty of Informatics and Information Technologies
Slovak University of Technology in Bratislava
Bratislava, Slovakia
{xvalastin, kristian.kostal}@stuba.sk

Abstract—The car-sharing market is constantly growing and Unfortunately, providing this kind of service to users brings
recently it has become even more popular than car ownership. a lot of security flaws; first of all is overall stability of service.
However, classic car-sharing system is based on a centralized Using DDoS attacks on centralized server would lead into
database server which can often lead to hacker attacks or pass-
word leaks. Moreover, in a classic car-sharing system, the owners collapse of the whole service. In comparison, when data are
of the cars can misuse customers’ data. As seen nowadays from distributed and processed over whole network of nodes like
a lot of use cases, the best solution to these problematic issues blockchain, it is practically impossible to tamper the service
is to use blockchain technology. Blockchain as decentralized, [5].
immutable, public ledger provides the customers with security In addition, traditional data servers provide us CRUD inter-
that is impossible to tamper. The aim of the proposed solution
is to create and implement peer-to-peer short term car-sharing face so we are able to create, read, update and delete data. This
application based on blockchain technology and smart contracts. suggests that some people might change users’ data without
For the implementation of smart contracts, Solidity programming further notification of the users. On the contrary, blockchain
language is used. Solidity works with Ethereum blockchain. The as immutable, public ledger allows us only to create and read
key novelty of the article is introducing a peer-to-peer car sharing data. If we want to update some information in blockchain, we
service without a central authority, what reflects a decrease of
costs and increase of data transparency in that system. Also token have to notify every single user in the network that we have
based solution gives us ability to cover business-to-business (B2B) changed data. As a matter of course, it is also impossible
and business-to-customer (B2C) use cases. to delete data from blockchain. It can be concluded that
Keywords—Blockchain; Car sharing; Asset sharing; Ethereum; blockchain as unchangeable method increases transparency
ERC-721; Smart contract and ensures data protection [6]. Moreover, using public ledger
allows us to cut fees which are needed for paying car sharing
I. I NTRODUCTION employees, such as system administrators.
An increasingly popular approach is the sharing of cars. All of these points mentioned above have become the core
While many people would not like to share a car with a motivation to create a decentralized peer-to-peer application
foreigner, car renting solutions such as car2go, where compa- built on top of Ethereum blockchain by using Solidity lan-
nies share they owned cars, are a great opportunity in cities. guage for smart contracts and ReactJS for client side of the
Gas powered vehicles are the main source of pollution and application. As a connector to Ethereum network we are
therefore the trend in many countries is to reduce the number using Web3. However, as mentioned before, we were not the
of cars on roads [1]. Car-sharing can solve also this problem first who had the idea to implement peer-to-peer car-sharing
by providing us with the car when we need it. According to platform. There are projects such as HireGo [7], DAV [8] or
statistic car is parked for 96% of the time [2]. Furthermore, Helbiz [9], but most of these project are in a stage of very
we are not forced to buy a car what significantly reduces early prototype.
our living and maintenance costs. Smart ideas side by side The following sections of this article are organized as
with mobility are calling for unconventional, probably even follows: Section II shows current state of the art solutions in
unfriendly advances [3]. That is why we want to propose a domain of car-sharing with blockchain. Section III describes
car sharing platform using blockchain technology. This idea design architecture of our decentralized car-sharing project
is not new and we are not the first ones to write about it. We with focus on our tokenized architecture. Evaluation and
came to it when we were thinking about possible use cases of Testing to proof our solution are in Section V. Section VI
the blockchain in real environment. When looking for ways sums up the article with results and we also discuss future
to share vehicles online, we found that many large companies work there.
and even startups are already working on specific solutions.
There are even startups specialized in this particular problem. II. R ELATED W ORK
One of the biggest peer-to-peer car-sharing platforms is In this section we are covering known solutions which are
Getaround [4]. That platform allows users to rent a car for for now considered state of the art.
a small period of time, starting from one hour. Users are able First application is HireGo. It is built on Ethereum using
to earn some money from sharing their cars when they do not smart contracts as a distributor of virtual fungible tokens
need them. (standard ERC-20) [10], called HGO. When we want to rent
978-1-7281-2181-9/19/$31.00 2019
c IEEE
 Figure 2. Demo application for scooter rental by HBZ

Figure 1. Demo application of HireGo

a car using HireGo, we need to exchange ETH to HGO

tokens. HireGo’s market place consists of three contracts. First
contract provides HGO tokens, second one provides ERC-721
car tokens and the last one provides rental contract. This rental
contract acts as an escrow and it is responsible for Vehicle
token between two parties [7]. In other words, when we rent Figure 3. Demo application for drone management
a car, we send HGO tokens to rental smart contract which
locks car token and HGO tokens for selected period. Then
the contract manages the transfer of the vehicle token from environment and also can fulfill user’s commands. As we see
owner to us. The rental solution also brings big security issue. in the Figure 3 with demo application we can simply interact
If we rent a car for a day trip and then we loose the phone with drone.
with Ethereum address, the car will be stuck for the rest of In contrast to other projects FFQuest’s main strength is
time and the owner is unable to help us because he/she does usage of their own chain called Distributed FFQ Ledger
not own that vehicle at the time of renting. In the Figure 1 based on Ethereum Virtual Machine. This ledger will pass
we can see alpha version of decentralized application made all transaction details between car companies, drivers and
by HireGo. Unfortunately, the developers allowed us only the customers such as payment conditions, assets, availability,
ability to create car for rent in this version. and reservations by using their own utility token, type ERC-
Next application we will introduce is Helbiz. Helbiz’s 20 called FFQ. This project uses central backend to mirror
philosophy is to create and integrate transport ecosystem, Distributed FFQ Ledger on a server and also for storing videos
where we are rewarded for using this system. This project is and images [11].
using well-known implementation of blockchain car-sharing The last known project is WONO. WONO used different
platform by providing ERC-20 tokens called HBZ. In contrast approach than the other projects, by providing a decentralized
to HireGo, we can also earn tokens by providing our data about market for renting services such as houses, cars and work-
driving to insurance companies [9]. Helbiz is the only project ing man days. Their solution is based on the set of smart
which has created a mobile application for renting electric contracts that work as a brigde between Ethereum main net
scooters as we see in the Figure 2. When we want to rent a and their blockchain. WONO blockchain is using Proof-of-
scooter we simply find a free one in the application and then Stake algorithm to ensure consensus in the network. They store
we scan QR code, and pay for the service with HBZ tokens. user’s profile data, represented as JSON, in the decentralized
If we do not have enough tokens, the app offers to buy HBZ IPFS file system. Data properties which contain confidential
tokens by using credit card. Unfortunately, we are not able information are encrypted on a client’s app side by a public
to test this solution because all of the scooters are located in key of the User’s address and can be decrypted only by private
Italy. key [12].
Next project we will look at is DAV. As opposed to other
projects the main goal is to connect autonomous vehicles (cars, III. D ESIGN
trucks and drones) with users. For communication purposes we As we mentioned earlier, there are some projects trying
use DAV token, which also serves as access token. Services to implement peer-to-peer car-sharing platform. Thus, we
in DAV network can be bought with virtual tokens based on will attempt to compare our project with those solutions’
Ethereum. It shall not be mistaken with DAV token. If we specifications. In contrast to related work and what we have
are owning vehicle or charging station which we provide to learnt, we try to build our solution with two different tokens
other users in network we are rewarded with DAV tokens [8]. based on non-fungible token type ERC-721 and one fungible
Autonomous vehicles in network automatically interact with token type ERC-20. We can simply describe ERC-721 [13]
 to understand token solution better.
The first scenario is client-to-client solution. The user
provides a car for renting. That means he is the owner of
unique non-fungible (NF) car token and also owner of the
NF master unlock token that could unlock the car anytime.
When the other user wants to rent a car for a period of time,
he will receive other NF unlock token for the particular car.
The metadata in unlock token describes details about renting.
When car owner decides to sell the car, car token and master
unlock token just simply change the owner. Also car can have
more master unlock tokens (something like spare car keys).
The second scenario describes usage of this kind of de-
Figure 4. Graphic representation of our solution sign for business-to-customer (B2C) and business-to-business
(B2B) solution. Our platform can also provide solution for
business car rental services such as ZipCar. Business service
will provide car fleet, which can be rented by users and other
as a smart contract which holds list of unique assets. These businesses. Then, for each car many master unlock tokens can
assets are represented as non-fungible tokens. First token type be created (e.g. for CEO, for person responsible for repairs,
represents car asset and the second one represents Unlock etc.), and also many rental unlock tokens can be created
token for car token. This solution will allow us to scale the (e.g. company rents a car for a long period of time to many
app more efficiently. The car owner, by the very nature of employees).
blockchain, would have full control over who gets access to
his or her information, and the same goes for the person IV. T EST SCENARIOS
borrowing the vehicle. Without a key to piece together the After implementation phase we wanted to make sure that
distributed data and then decrypt it, it is far harder for a our application is working, so we created two test scenarios.
third party to hack either side’s personal information. We The main goal was to test overall speed, stability and func-
applied ERC-20 token to reward user for using our car-sharing tionality of our proposed solution.
application. After user collects particular amount of tokens, Our testing environment consisted of a MacBook Pro with
he can apply for a discount to renting a vehicle. Users are MacOS Mojave (10.14.3), Chrome web browser with enabled
then more motivated to use our service if they can get it JavaScript and installed MetaMask plugin. MetaMask provides
cheaper. Also we managed to save car token’s image to IPFS, simple interface for communication with ethereum network.
which is decentralized storage, to reduce costs of ethereum Also for testing purposes we did not deploy smart contract
transactions. Storing IPFS URI is much more effective and to main or any test network. Instead we created our local
cheaper than storing image on the blockchain network. As a ethereum network using Ganache. Ganache is a one click
gateway to IPFS network, we decided to use Pinata instead of personal Ethereum blockchain.
hosting our own, because of really simple implementation to
JavaScript, fair pricing and overall speed of service [14]. At V. R ESULTS
the beginning of this project we were thinking how to provide First scenario covers creating new car available for renting
as wide scalability as possible. Moreover, we bore in mind and also creating unlock token for the user. Every transaction
that providing users positive experience will convince other on blockchain network needs to be payed by user. The motiva-
users to use our application. In the Figure 4 we can see the tion behind this test was to get actual price that user needs to
user interaction with the system. When we want to rent a car, pay in order to create a car available for renting. Also we can
we just select the car, length of the renting period, and pay consider this amount of money as initial investment for user to
with our preferred currency (USD/EUR). Application in the list the car. The process of creating new car is pretty straight-
background exchanges money to cryptocurrency (ETH in our forward. We simply open web dApp, log in to the network via
case) and after that executed smart contract will provide us MetaMask, fill out the form with name, description and picture
Unlock Token to unlock the car. In our project, we do not of the car. The process is simple, user friendly and the whole
force users to buy ETH and convert them into other fungible scenario takes only 15 seconds to complete. Calling create car
tokens. Instead we automatically convert real currency to ETH token function costs user about 0.014485 ETH.
exactly in time when user needs it. In second scenario we tested a time and responsiveness of
We use an IoT device in a car, which is waiting for unlock our application. Our test covers case from opening application
token. The IoT device can be managed by any standard means to open details with car and receive token for unlocking. We
or also by blockchain as introduced e.g. in [15], [16]. If vehicle wanted to find out how long does it rent and unlock car. This
receives unlock token, it will verify its validity with blockchain test scenario is crucial for our application, because one of our
network. After validation being successful, the car is ready to motivation is to provide smooth user experience. If the test
be taken for a ride. We take on two scenarios which help us scenario would take too long, user can become impatient and
 TABLE I. Test results of time needed for renting
ACKNOWLEDGMENT
Test nr. Time (s) This research was supported by the Ministry of Education,
Test 1 13.41 Science, Research and Sport of the Slovak Republic, Incentives
Test 2 11.11 for Research and Development, grant agreement number:
Test 3 10.35 2018/14427:1-26C0. It is also a part of APVV-15-0731 project.
Test 4 12.9
Test 5 14.78 R EFERENCES
Test 6 10.40 [1] D. Rojas-Rueda, A. de Nazelle, O. Teixidó, and M. Nieuwenhuijsen,
Test 7 12.73 “Replacing car trips by increasing bike and public transport in the
greater Barcelona metropolitan area: A health impact assessment study,”
Test 8 13.30 Environment International, vol. 49, pp. 100–109, nov 2012.
Test 9 11.57 [2] RAC Foundation, “Spaced Out parking report.” Accessed: 2019-04-15.
Test 10 10.91 [3] P. K. Sharma, S. Y. Moon, and J. H. Park, “Block-VN: A Distributed
Blockchain Based Vehicular Network Architecture in Smart City,”
Average 12.146 Journal of Information Processing Systems, vol. 13, no. 1, pp. 184–195,
2017.
[4] Getaround, “Getaround - Peer-to-peer car sharing and local car rental.”
Accessed: 2019-02-24.
could leave the app very soon. However, based on results, [5] S. Nakamoto, “Bitcoin: A Peer-to-Peer Electronic Cash System,” tech.
rep., 2008.
using the application for renting a car is fast and comfortable. [6] X. Xu, C. Pautasso, L. Zhu, V. Gramoli, A. Ponomarev, A. B. Tran,
As we see in the Table I, the average time for purchasing a and S. Chen, “The Blockchain as a Software Connector,” in 2016
rent token is 12.146 s. 13th Working IEEE/IFIP Conference on Software Architecture (WICSA),
pp. 182–191, IEEE, apr 2016.
[7] HireGo, “HireGo - Decentralised Shared Mobility Platform,” tech. rep.,
2018.
VI. C ONCLUSIONS AND F UTURE W ORK [8] T. Copel, Noam and Ater, “DAV White Paper,” tech. rep., 2017.
[9] Helbiz, “Helbiz Mobility System.” Accessed: 2018-12-07.
[10] F. Vogelsteller and V. Buterin, “Erc-20 token standard,” Ethereum
Our primary focus was to create a peer-to-peer car-sharing Foundation (Stiftung Ethereum), Zug, Switzerland, 2015.
[11] FFQuest, “FFQuest - Blockchain marketplace for Car rental, Ride-
decentralized application. We were able to fix problems that sharing and Parking.” Accessed: 2019-04-15.
centralized services were facing such as denial of service [12] WONO, “WONO White Paper,” tech. rep., 2018.
and data mutability by providing decentralized and immutable [13] W. Entriken, D. Shirley, J. Evans, and N. Sachs, “Non-Fungible Token
Standard, document ERC-721, Sep. 2018.” Accessed: 2018-12-07.
blockchain technology. Our solution adds security to the [14] “Pinata - Add Files To IPFS Effortlessly.” Accessed: 2019-04-15.
whole ecosystem, firstly by utilizing blockchain, and secondly [15] P. Helebrandt, M. Belluš, M. Ries, I. Kotuliak, and V. Khilenko,
by using two different types of ERC-721. Again, using the “Blockchain Adoption for Monitoring and Management of Enterprise
Networks,” in 2018 9th IEEE Annual Information Technology, Elec-
example of hijacking a car, blockchain can make it much tronics and Mobile Communication Conference (IEMCON), (Vancouver,
harder for a hacker to gain control, because it is only being Canada), pp. 1–5, IEEE, 2018.
given a digital "key" to access the vehicle’s systems. If this [16] K. Košt’ál, P. Helebrandt, M. Belluš, M. Ries, and I. Kotuliak, “Man-
agement and Monitoring of IoT Devices Using Blockchain,” Sensors,
key is distributed around the internet, and accessible only with vol. 19, p. 856, feb 2019.
its owner’s permission, it is immeasurably harder to hack than
a cybercriminal being faced with only hacking their way into
a single server. Tokenization of car keys helped us to cover
many use cases connected with car-sharing services, such as
preventing car from partial deadlock. As we mentioned be-
fore, current implementation of our decentralized application
covers major scenarios such as creating the car token with
major unlock token and providing tokenized renting solution.
According to test scenarios, creating a car available for renting
seems to be pretty straightforward. Additionally, users are able
to rent a car in average 12.146 seconds, which brings the same
user experience as using centralized car-sharing solution.
In near future we are going to add different Unlock token
types, because actual implementation allows us only to create
master unlock token. Moreover, our design proposed automatic
exchange currency to cryptocurrency, but we did not find
simple exchange market for this kind of use case. Additional
improvement proposal is to provide seamless user experience
with log-in to our application which means that users are not
forced to remember their private keys and they will be able to
use their email address instead of it.

View publication stats