Towards Developing an Ontology for a Digital Twin in Battery Testing

Nuno Marques

1 a

, Marco Rodrigues

1 b

, Mannin Himanshu

2 c

and Foad Gandoman

2 d

INEGI - Institute of Science and Innovation in Mechanical and Industrial Engineering, Porto, Portugal

RSTER - Reliability & Safety Technical Centre, Brussels, Belgium

Keywords:

Digital Twin, Battery Testing, Ontology, OWL, DTDL, Azure Digital Twins.

Abstract:

Ontologies are of great importance in organizing and structuring domain knowledge towards interoperability

and data integration in various ﬁelds, as provide a standarised vocabulary and a formal representation of

relationships between concepts, which is essential for advancing data-driven applications. This paper presents

the development of an OWL (Web Ontology Language) ontology for the battery testing ﬁeld, creating the

foundations for the development of a Digital Twin that will virtualize tests to be performed on battery cells

and modules. It emphasizes the signiﬁcance of the battery domain characterization and ontology deﬁnition

as critical components in developing an effective Digital Twin for battery testing. An investigation of prior

studies available in the literature was conducted, highlighting examples of ontologies such as SSN (Semantic

Sensor Network) and SOSA (Sensor, Observation, Sample, and Actuator), which targets integration into the

digital twin environments to enhance sensor data management and interoperability. The research also found

hybrid ontologies, combining elements from existing ones and battery-speciﬁc Digital Twin architectures. The

developed ontology was validated through a practical use-case by integration with cloud platform Microsoft

Azure Digital Twins, converting the ontology from OWL to DTDL (Digital Twins Deﬁnition Language).

This step completes the cycle as the proposed framework aims to create a robust and scalable Digital Twin

environment that can be adapted to various battery testing scenarios, providing actionable insights from tests.

1 INTRODUCTION

The incorporation of Digital Twins in sophisticated

battery technologies offers a revolutionary method

for testing and optimizing battery performance in

the ever-changing landscape of battery technol-

ogy (Naseri et al., 2023). A Digital Twin depicts

a physical system, allowing for continuous monitor-

ing, simulation and data analysis. When it comes

to battery testing, a Digital Twin can greatly expand

our comprehension of battery performance in differ-

ent situations, thereby enhancing efﬁciency, reliabil-

ity, and lifespan (Javaid et al., 2023).

The development of a Digital Twin for the battery

testing domain involves several critical steps, starting

with the deﬁnition of the ontology. This, establishes a

structured framework for organizing and interpreting

the diverse parameters and data ﬁelds associated with

battery testing providing a vocabulary and set of rela-

https://orcid.org/0009-0007-1136-9865

https://orcid.org/0000-0002-6387-0105

https://orcid.org/0000-0002-3749-0125

https://orcid.org/0000-0003-3814-4547

tionships that describe the characteristics, behaviors,

and interactions of the various elements within the

battery tests. This standardized framework ensures

consistency and interoperability across different data

sources and analytical tools (Karabulut et al., 2024).

The groundwork needed to characterize the on-

tology involves the identiﬁcation and cataloging of

all relevant data assets required for the Digital Twin.

This includes sensors, metadata, simulation models,

among others. By mapping these, we created a de-

tailed blueprint that outlines how data ﬂows between

different components of the Digital Twin, facilitat-

ing integration and efﬁcient data management (Haller

et al., 2017). This introduction sets the stage for

a deeper exploration of these concepts, highlighting

their importance in creating a reliable Digital Twin

that not only mirrors the physical battery system but

also provides actionable insights for optimizing bat-

tery tests.

Marques, N., Rodrigues, M., Himanshu, M. and Gandoman, F.

Towards Developing an Ontology for a Digital Twin in Battery Testing.

DOI: 10.5220/0013082300003838

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 16th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2024) - Volume 2: KEOD, pages 279-286

ISBN: 978-989-758-716-0; ISSN: 2184-3228

279

2 ONTOLOGIES IN DIGITAL

TWIN DEVELOPMENT

2.1 Overview

The concept of Digital Twins represents a signiﬁcant

advancement in the ﬁeld of batteries testing and devel-

opment. By constructing a virtual replica, researchers

will have the capability to virtualize a range of sce-

narios, thereby getting faster results. This chapter ex-

plores speciﬁc ontologies that support Digital Twin

technologies within the context of battery systems. It

discusses their current state and examines potential

modiﬁcations that could facilitate more effective vir-

tual testing of batteries. The adaptability and exten-

sion of these ontologies are important for the accurate

representation and analysis of battery behaviors un-

der various test conditions, thus contributing to more

precise and reliable battery technology developments.

Digital Twin technologies for batteries may need

the integration of complex relationships and rules de-

rived from multiple data sources. The implementation

of well-deﬁned ontologies in these systems is crucial

for achieving smooth interoperability and consistency

on the designed data model. This approach ensures

that the Digital Twins accurately reﬂect their real-

world counterparts and provide meaningful insights

that are essential for battery testing scenarios.

2.2 Literature Review

Originally developed for sensor data integration and

IoT applications, the SSN (Semantic Sensor Network)

and SOSA (Sensor, Observation, Sample, and Actua-

tor) ontologies provide frameworks that can be partic-

ularly beneﬁcial for batteries Digital Twins (Karabu-

lut et al., 2024; Janowicz et al., 2019), as they de-

scribe sensors and measurements, crucial for moni-

toring battery conditions. (Bamunuarachchi et al.,

2021). This integration allows for a more detailed

and structured representation of sensor data, observa-

tions, and actions within the system, facilitating im-

proved interoperability and data analysis capabilities.

This dual ontology approach leverages the strengths

of both to provide a comprehensive framework that

supports advanced monitoring and control function-

alities within the Digital Twin environment.

The IoT ontologies, such as those proposed on

Digital Twins in cyber-physical systems, offer robust

structures for integrating real-time data into Digital

Twin environments (Steinmetz et al., 2018). These

ontologies are designed to handle dynamic and com-

plex data streams, which are common in battery op-

eration and testing scenarios. Implementing these

IoT ontologies in battery Digital Twins would fa-

cilitate the integration of sensor data, enhancing the

Twin’s ability to simulate and predict battery behav-

ior under different conditions. The work by (Merkle

et al., 2019) outlines an architecture for a Digital Twin

speciﬁcally designed for battery systems. This archi-

tecture could guide the development of a specialized

ontology that addresses the unique needs of battery

testing, such as lifecycle management, degradation

modeling and performance optimization.

Other usages of ontologies are found to describe

physical components, their physical attributes, states

in a system and the relation in between them use an

ontology to describe physical parts of a plant, such as

root, stem or leaf, and the ontology is then used to ex-

tract rules for decision making (Skobelev et al., 2020).

An example in manufacturing is one using an ontol-

ogy for CNC (Computer Numerical Control) machine

tool that includes concepts such as Material, Person-

nel, Device, and Environment (Liu et al., 2020).

Recent research demonstrates the growing use

of ontologies for integrating and managing com-

plex data in various domains. A relevant exam-

ple is one developed to address sensor uncertainties

in autonomous vehicles, enhancing safety and reli-

ability by structuring knowledge about sensor limi-

tations and environmental impacts (Alharbi. and A.

Karimi., 2023). Other, to improve data synchro-

nization between SysML models and heterogeneous

data sources, promoting consistency in systems engi-

neering (Zhang. et al., 2023). These studies, along

with others that apply ontologies to physical compo-

nents and manufacturing processes (Skobelev et al.,

2020; Liu et al., 2020), highlight the potential for

a hybrid ontology combining IoT frameworks and

battery-speciﬁc Digital Twin architectures to support

advanced battery management.

For the speciﬁc needs of virtualizing battery test-

ing, a hybrid ontology combining elements from both

SSN/SOSA for sensor data handling and speciﬁc con-

structs from battery-focused Digital Twin architec-

tures could be appropriate. It should include deﬁni-

tions of battery cells and modules, including physical

and chemical properties, standardized descriptors for

test conditions, procedures and expected outcomes,

and descriptions of performance metrics such as en-

ergy capacity, charge/discharge rates, and efﬁciency.

3 BATTERIES

CHARACTERIZATION

This work considered battery testing for three distinct

use cases: Offroad and Industrial, Automotive and

DTO 2024 - Special Session on Ontologies for Digital Twin

280

Stationary applications. Each has its own distinct par-

ticularities and technical characteristics, therefore, it

is relevant for the Digital Twin to specify the baseline

for each use-case typical cell and module. Further-

more, each use case has unique requirements, which

may result in different tests sequences or procedures.

Additionally, this characterization boosts traceability,

helping in the identiﬁcation of faults speciﬁc to each

use case and in compiling useful statistics. Such de-

tailed tracking will enhance the capabilities of the

Digital Twin, providing a more complete solution and

opening doors for future and more effective analytics.

3.1 Offroad and Industrial Devices

Table 1 presents cell technical speciﬁcations and Ta-

ble 2 summarizes at module level, for the Offroad and

Industrial use-case.

Table 1: Cell Level Data.

Parameters Description Value

Capacity 280 Ah

Voltage 3,2V

Impedance

Resis-

tance(1KHz)

≤0.25 mΩ

Standard

charge and

discharge

Charge/discharge cur-

rent

0.5 C/0.5 C

Standard

charge and

discharge

Cut off voltage of

charge/discharge

3.65V/2.5V

Maximum

charge

/discharge

current

Continuous

charge/discharge

1C/1C

Maximum

charge

/discharge

current

Pulse

charge/discharge

(30s)

2C/2C

Charging

Temp.

0°C∼55 °C

Discharging

Temp.

-20°C∼55 °C

Chemistry/

Cell Type

Prismatic cell

LIFEPO4

LFP

Table 2: Module Level Data.

Parameters Value

Capacity 560 Ah

Voltage 6.4V

Conﬁguration 2S2P

SOC Range 70-80

Table 3 displays measurements related to the bat-

tery cell testing procedure carried out during the in-

coming quality control phase. This step is performed

to verify the compliance of the item with the manu-

facturer’s datasheet and to collect information needed

for predictive maintenance models. This information

may serve to the Digital Twin as normal baseline val-

ues.

Table 3: Cell Testing Data in Control Phase.

Parameters Value

Resistance Pre-Charge 0.35 mΩ

Resistance Post-Charge 0.15 mΩ

Capacity Residual 218.01 Ah

Capacity Tot 621 Ah

Impedance 0.105 mΩ

OCV V 3.295

3.2 Automotive Devices

Table 4 presents essential technical requirements for

batteries intended for use in the automotive sector.

These speciﬁcations are crucial to ensure that the bat-

teries not only meet performance and safety expecta-

tions but also facilitate effective integration with ve-

hicle electrical systems.

The criteria listed include the operational voltage

range, which signiﬁcantly varies from 90V to 450V

to accommodate different types of vehicles. Addition-

ally, a usable energy capacity ranging from 30 kWh to

100 kWh is speciﬁed, which is fundamental in deter-

mining the vehicle’s range under various driving con-

ditions.

Another key feature outlined in the table is the

rapid charging performance, allowing the battery to

reach 80% capacity in just 20 minutes—a critical

characteristic for market acceptance of electric vehi-

cles. These parameters not only reﬂect the direct tech-

nical needs of electric vehicles but also highlight the

challenges associated with designing and testing bat-

tery systems that must be robust, efﬁcient, and capa-

ble of handling the dynamic demands of modern elec-

tric mobility.

3.3 Stationary

Stationary or Battery Energy Storage Systems

(BESS) considered for this work use a modulariza-

tion approach where multiple modules are intercon-

nected to reach the desired energy requirements. Cur-

rently it uses a minimum of three modules which can

deliver usable energy up to 7.5 kWh. It can be fur-

ther extended output up to 15 kWh by connecting

seven modules. The modules utilize cylindrical cells

of 18650 format and can be operated in a wide range

of temperature (-15 and 55°C). Table 5 and Table 6

shows the cell and module level data speciﬁcations,

Towards Developing an Ontology for a Digital Twin in Battery Testing

281

Table 4: Cell and Module Data.

Parameter Value

Chemistry LFP

Nominal operation volt-

age range

90V-450V

Usable Total Energy Ca-

pacity

30 kWh - 100 kWh

SOC of operation range 0%-80%

Fast charging perfor-

mance

0%-80% in 20min./1.7C

Peak discharge power 200kW

Temperature operating

range

Charge: -20°C and 55°C

Discharge: -30°C and

55°C

Storage temperature

range

-40°C ≤ 60°C

Lifecycle of application 160.000 km 1100 full cy-

cles

Battery Pack Weight

(1pack)

≤599kg

Energy Density - Pack ≥201Wh/kg

Energy Density - Mod-

ule

≥281Wh/kg

Power Density – Pack

(1C)

≥201Wh/kg

Power Density – Module

(1C)

≥281Wh/kg

respectively.

Table 5: Cell Level Data.

Parameter Value

Cell model TerraE 30E

Cell chemistry Li-ion NMC

Format 18650 Cylindrical

Voltage range 3.2∼4.08V

4 ONTOLOGY DEVELOPMENT

FOR BATTERY TESTING

After collecting all relevant data so far for the de-

tailed modelling of the problem within the battery

testing domain, the next step was to create the ontol-

ogy. For that, Prot

e was used (Stanford Center for

Biomedical Informatics Research, 2016), which is a

comprehensive, free, open-source ontology editor and

knowledge management system. It supports various

ontology languages, including Web Ontology Lan-

guage (OWL), facilitating the creation, manipulation

and sharing of complex knowledge structures. It also

features a user-friendly graphical interface that allows

for design, visualization, and testing of ontologies,

which makes it an ideal tool for modeling detailed

ontological structures that deﬁne the relationships and

properties of entities within speciﬁc domains.

In order to create an ontology ((Marques, 2024),

Table 6: Module Level Data.

Parameter Value

Nominal Voltage 51.4V

Nominal energy 3.3 kWh

Usable energy 2.5 kWh

Charging current (max.) 29 A

Discharging current

(cont.)

30 A

Discharging current peak

(3s)

40 A

Depth of discharge

(DoD)

≤79%

Charge operating tem-

perature range

0 and 45ºC

Discharge operating

temperature range

-15 and 55ºC

Internal protection of

casing

IP20

Certiﬁcation IEC62619/VDE-AR-

E2510-50/CE/UN38.3

four components are essential: Classes, Relation-

ships, Properties and instances (also known as In-

dividuals). In the following sub-section, these four

components will be explained and detailed.

4.1 Class

Being a Class, in the context of ontology design, a

fundamental concept that represents a set of objects

or instances sharing common characteristics or at-

tributes, within this work, the following Classes were

identiﬁed:

1. Components;

a. Cell;

b. Module;

2. Test;

3. Test Bench;

4. Test Procedures

Components are the entities that will be tested, and

Cell and Module are a subclass of Components. The

class Test, as the name suggests, was created to char-

acterize the test that will be performed on a Compo-

nent. The same logic applies to the classes Test Bench

and Test Procedures.

4.2 Properties

Properties describe attributes or characteristics of the

classes, deﬁning speciﬁc aspects of the class in-

stances, such as measurements, conditions, or de-

scriptive elements as well as their data types, enabling

a more detailed and structured representation of the

data within the ontology. Figure 1 displays all the

DTO 2024 - Special Session on Ontologies for Digital Twin

282

properties relevant to each class considered within

this work. Since both Cell and Module are subclasses

of the class Component, they share all data properties

present in the Component class in addition to each

speciﬁc properties. The Test class contains two main

data categories: test properties and test telemetry. Test

properties are important for deﬁning a test with pa-

rameters such as Test Type and Test Procedure ID.

Test telemetry considers parameters such as Time and

Current, being responsible for recording the outputs

of each test. Telemetry data is the only dynamic data

among the properties.

Figure 1: Properties of each class.

4.3 Relationships

Relationships describe how classes and instances are

connected, deﬁning the associations and dependen-

cies between entities and helping to create a struc-

tured and meaningful representation of the data and

comprehensive understanding of the domain. Within

this work, the following relationships have been iden-

tiﬁed: Belong, that indicates that one entity is a part of

another, Contains, shows that one entity includes an-

other, Made in, speciﬁes the location or environment

where an activity takes place and Made to, deﬁnes

the target or subject of an activity.

4.4 Constraints

Constraints specify conditions that must be met for a

property (such as object or data properties) or a re-

lationship to hold true for a particular ontological el-

ement, such as cardinality restrictions, which deter-

mine how many times a property can be used, and

value restrictions, which deﬁne the types of values a

property can accept. Together, constraints ensure that

relationships between entities within the ontology re-

main coherent and logical. In our ontology, the fol-

lowing constraints were designed:

• Constraint 1 (Test). Each Test must include one

Component (Cell or Module), one Test Procedure

and one Test Bench;

• Constraint 2 (Components). Each Component

must be classiﬁed wither as Cell or Module;

Since OWL is a declarative language, it doesn’t allow

conditional logic, so other potential constraints have

been left out for this phase of the work, such as ones

specifying ranges for data properties depending on the

use-case, or depending on the test performed.

4.5 Instantiation and Validation

Instances represent concrete application examples of

the classes deﬁned in the ontology. They give real

case scenarios to the abstract concepts and relation-

ships. By creating instances, the accuracy and con-

sistency of the ontology can be tested, ensuring that

the deﬁned classes, properties and relationships will

model the real-world entities and their interactions.

As an example, an instance of the class ”Cell” might

be a speciﬁc type of cell used in a battery test, while

an instance of the class Test could be a particular test

conducted on a cell. This validates the ontology’s

structure and functionality, verifying that all neces-

sary elements are correctly represented and intercon-

nected. Figure 2 displays an instance of the class

”Cell” named ”Cell 1”.

Figure 2: Instance of class ”Cell”.

As shown in Figure 2, the properties attributed to

the cell class have been populated with real values

creating an instance of the class ”Cell”.

By creating Instances, Prot

e also allows to or-

ganize the created structures on a schematic manner,

as seen in Figure 3. Here, an instance of the class

”Module”, contains ”Cell 1” through ”Cell 4” that are

instances of the class ”Cell”.

Figure 4 illustrates ”Preconditioning Test 1”, as

Towards Developing an Ontology for a Digital Twin in Battery Testing

283

Figure 3: Graphical Representation of the Ontology: An

Instance of the ”Module” Class Named ”Module 1”.

an instance of the class Test, that belongs to the ”Test

Procedures 1”, an instance of the class ”Test Proce-

dures”, which was performed in the ”Test Bench 1”,

an instance of the Test Bench class and was made to

”Cell 1”, an instance of the class ”Cell”.

Figure 4: Representation of the Ontology: An instance of

the ’Test’ Class as ’Preconditioning Test 1’.

The instantiation of classes validates the ontology,

as it creates scenarios that test the model and ensures

its suitability for the domain it represents. The deﬁni-

tion of the ontology provides a structured and compre-

hensive framework for representing the battery test-

ing domain with the entities, properties, and relation-

ships.

5 FROM OWL ONTOLOGY TO A

CLOUD PLATFORM - A CASE

STUDY WITH MICROSOFT

AZURE DIGITAL TWINS

Integrating ontologies with cloud platforms such as

Microsoft Azure Digital Twins provides a scalable,

ﬂexible and interoperable environment for battery

systems. This chapter focuses on incorporating the

OWL ontology within Azure Digital Twins, empha-

sizing the process of converting the ontology from

OWL to Digital Twins Deﬁnition Language (DTDL),

which is the standard modeling language used by

Azure Digital Twins.

5.1 Azure Digital Twins and Digital

Twins Deﬁnition Language (DTDL)

Azure Digital Twins is a comprehensive cloud-based

platform designed for building digital replicas of

physical environments. It uses DTDL ((Marques,

2024)), which is a JSON-based modeling language

designed for deﬁning Digital Twin models, prop-

erties, telemetry and relationships in a machine-

readable format. The language supports the integra-

tion of diverse standards, enabling interoperability

across Digital Twin environments. Through DTDL

its possible to deﬁne complex models that accurately

represent battery components, such as cells and mod-

ules, and their respective properties, including volt-

age, temperature and state of charge. This provides a

robust framework for modelling complex systems and

enhancing their performance through real-time data

integration.

5.2 From OWL to DTDL

To integrate the developed ontology within Azure

Digital Twins, it had to be converted from the OWL

format to DTDL, using a manual process that in-

volved several key steps to ensure the Digital Twin

models remain accurate and consistent with the orig-

inal ontology. There are also available free tools to

help this process (Kevin Hilscher, 2020).

The initial step was to identify the core entities

and relationships in the OWL ontology that need to be

represented in DTDL. In this work, these were ”Cell,”

”Module,” ”Test,” ”Test Bench,” and ”Test Proce-

dures,” as well as their associated properties and re-

lationships, such as ”Contains,” ”Belong,” ”Made in,”

and ”Made to.” Each OWL class is translated into a

corresponding DTDL model. For instance, the ”Cell”

class in OWL becomes a DTDL model that speciﬁes

properties relevant to a battery cell, such as voltage

and temperature, formatted in a JSON-like structure

as shown in Figure 5.

In this convertion, properties deﬁned in the OWL

ontology, such as those for the ”Cell” class, were

mapped to equivalent DTDL properties or telemetry,

ensuring that each property retains its intended data

type and functionality. Relationships in the OWL on-

tology, like the one indicating a ”Module” contains

”Cells,” were represented in DTDL using the ”Re-

lationship” type, maintaining the logical connections

and dependencies established in the OWL ontology.

DTO 2024 - Special Session on Ontologies for Digital Twin

284

Figure 5: Example of JSON as part of the Test class in Dig-

ital Twins Deﬁnition Language (DTDL).

5.3 Deployment on Azure Digital Twins

Once the ontology was successfully converted to

DTDL, the next step was to deploy to the Azure Dig-

ital Twins platform. This process involved uploading

the DTDL models to the Azure Digital Twins instance

using the Azure portal or through Azure SDKs and

APIs. Upon deployment, these models form a digi-

tal representation of the battery testing system in the

cloud, closely aligned with the structure deﬁned in the

OWL ontology, as illustrated in Figure 6.

Figure 6: Ontology representation on Azure Digital Twins.

Subsequently, real-time data streams from IoT

sensors and devices can be integrated in Azure Dig-

ital Twins. These data streams can be mapped to their

corresponding DTDL telemetry properties, enabling

continuous monitoring and analysis of battery perfor-

mance across various test conditions. Additionally,

custom logic and workﬂows can be implemented us-

ing other Microsoft Azure resources, such as Azure

Functions, Logic Apps, or others, to automate actions,

generate alerts and conduct calculations based on the

Digital Twin data. For example, an alert could be set

to trigger if the temperature of a battery cell exceeds

a certain threshold, or a speciﬁc test procedure could

be initiated automatically.

5.4 Advantages of Cloud Integration

with Azure Digital Twins

Integrating the battery testing ontology with Azure

Digital Twins provides several beneﬁts. It enables

real-time monitoring and feedback, which allows for

proactive maintenance and more responsive decision-

making. By continuously ingesting data from IoT

sensors and physical devices, the Digital Twin envi-

ronment remains dynamically updated, reﬂecting the

latest conditions and states of the battery systems dur-

ing tests. Furthermore, the use of Azure’s advanced

analytics tools may enhance the predictive capabili-

ties of the Digital Twins, allowing more precise pre-

dictions about battery degradation patterns, perfor-

mance optimization opportunities and potential fail-

ures, thus, enabling more efﬁcient and targeted inter-

ventions during the testing stages. Additionally, the

scalable infrastructure of Azure ensures that the digi-

tal twin environment can handle increasingly complex

and large-scale battery testing systems. This scalabil-

ity makes the solution both cost-effective and future-

proof, accommodating growth and evolution in bat-

tery technology and testing requirements.

6 CONCLUSIONS AND FUTURE

WORK

Ontologies are crucial in Knowledge Engineering as

they provide a structured framework for represent-

ing and organizing knowledge in a speciﬁc domain,

contributing to the standardization and understand-

ing across different systems, applications, and stake-

holders, ensuring consistency and reducing ambigu-

ity. Ontologies are the ﬁrst step towards interoper-

ability between different systems and organizations,

driving to an effective and uniform data interpretation

among all. This work has enabled the development of

an Ontology in a standardised format for the ﬁeld of

battery testing. The literature review identiﬁed some

Ontologies aimed at Digital Twins development, but

none for the area of battery testing. As such, this

work will ﬁll the gap identiﬁed. The development of

the Ontology using a standardised format makes it ag-

nostic and applicable to different use-cases. Further-

more, practical steps were taken to validate the appli-

cability and ﬂexibility of the ontology, through cre-

ating instances and integrating with Microsoft Azure

Digital Twins platform. The conversion from OWL

to DTDL leverages Azure’s cloud capabilities. This

framework facilitates data integration and manage-

ment, from raw sensor data to user-selected tests, en-

suring efﬁcient communication between physical and

Towards Developing an Ontology for a Digital Twin in Battery Testing

285

virtual testing environments. This cloud-based Digi-

tal Twin will continue to support ongoing innovation

in battery technology, providing a ﬂexible platform,

adaptable to various contexts and technological envi-

ronments.

This research establishes a foundational ontology

for Digital Twins in battery testing. Nevertheless, be-

ing this an ongoing work, it is expected that future ef-

forts will focus on consolidating the developed ontol-

ogy, through concrete speciﬁcation of all Tests, Test

Procedures or other domain speciﬁcations not identi-

ﬁed yet. This groundwork will allow further integra-

tion of real-time data streams from physical battery

test benches, ensuring interoperability and validating

the applicability of the ontology via Microsoft Azure

Digital Twins cloud platform. These advancements

will help create a more robust, adaptable, and com-

prehensive Digital Twin environment for battery test-

ing, driving innovation and optimization in this criti-

cal ﬁeld.

ACKNOWLEDGEMENTS

Funded by the European Union: Horizon Europe re-

search and innovation programme under Grant Agree-

ment No. 101103755 (FASTEST: Fast-track hybrid

testing platform for the development of battery sys-

tems). Views and opinions expressed are however

those of the author(s) only and do not necessarily re-

ﬂect those of the European Union or CINEA. Neither

the European Union nor the granting authority can be

held responsible for them.

The authors acknowledge Fundac¸

ao para a

encia e a Tecnologia (FCT) for its ﬁnancial sup-

port via the project UIDB/50022/2020 (LAETA Base

Funding).

REFERENCES

Alharbi., M. and A. Karimi., H. (2023). Towards develop-

ing an ontology for safety of navigation sensors in au-

tonomous vehicles. In Proceedings of the 15th Inter-

national Joint Conference on Knowledge Discovery,

Knowledge Engineering and Knowledge Management

- KEOD, pages 231–239. INSTICC, SciTePress.

Bamunuarachchi, D., Georgakopoulos, D., Jayaraman, P. P.,

and Banerjee, A. (2021). A framework for enabling

cyber-twins based industry 4.0 application develop-

ment. In 2021 IEEE International Conference on Ser-

vices Computing (SCC), pages 340–350.

Haller, A., Janowicz, K., Cox, S., Phuoc, D., Taylor, K.,

and Lefranc¸ois, M. (2017). Semantic sensor network

ontology.

Janowicz, K., Haller, A., Cox, S. J., Le Phuoc, D., and

Lefranc¸ois, M. (2019). Sosa: A lightweight ontol-

ogy for sensors, observations, samples, and actuators.

Journal of Web Semantics, 56:1–10.

Javaid, M., Haleem, A., and Suman, R. (2023). Digital twin

applications toward industry 4.0: A review. Cognitive

Robotics, 3:71–92.

Karabulut, E., Pileggi, S. F., Groth, P., and Degeler, V.

(2024). Ontologies in digital twins: A systematic lit-

erature review. Future Generation Computer Systems,

153:442–456.

Kevin Hilscher (2020). Converter to dtdl. https://github.

com/Azure-Samples/RdfToDtdlConverter.

Liu, J., Yu, D., Bi, X., Hu, Y., Yu, H., and Li, B. (2020).

The research of ontology-based digital twin machine

tool modeling. In 2020 IEEE 6th International Con-

ference on Computer and Communications (ICCC),

pages 2130–2134.

Marques, N. (2024). Ontology for battery test-

ing. https://github.com/Ineginmarques/

Ineginmarques-Ontology-for-Battery-Testing.

Merkle, L., Segura, A. S., Torben Grummel, J., and

Lienkamp, M. (2019). Architecture of a digital twin

for enabling digital services for battery systems. In

2019 IEEE International Conference on Industrial

Cyber Physical Systems (ICPS), pages 155–160.

Naseri, F., Gil, S., Barbu, C., Cetkin, E., Yarimca, G.,

Jensen, A., Larsen, P., and Gomes, C. (2023). Digital

twin of electric vehicle battery systems: Comprehen-

sive review of the use cases, requirements, and plat-

forms. Renewable and Sustainable Energy Reviews,

179:113280.

Skobelev, P., Laryukhin, V., Simonova, E., Goryanin, O.,

Yalovenko, V., and Yalovenko, O. (2020). Multi-agent

approach for developing a digital twin of wheat. In

2020 IEEE International Conference on Smart Com-

puting (SMARTCOMP), pages 268–273.

Stanford Center for Biomedical Informatics Research

(2016). Prot

e. https://protege.stanford.edu/.

Steinmetz, C., Rettberg, A., Ribeiro, F. G. C., Schroeder,

G., and Pereira, C. E. (2018). Internet of things ontol-

ogy for digital twin in cyber physical systems. In 2018

VIII Brazilian Symposium on Computing Systems En-

gineering (SBESC), pages 154–159.

Zhang., Y., Jacobs., G., Zhao., J., Berroth., J., and

Hoepfner., G. (2023). Development of an owl ontol-

ogy based on the function-oriented system architec-

ture to support data synchronization between sysml

and domain models. In Proceedings of the 15th Inter-

national Joint Conference on Knowledge Discovery,

Knowledge Engineering and Knowledge Management

- KEOD, pages 143–154. INSTICC, SciTePress.

DTO 2024 - Special Session on Ontologies for Digital Twin

286