20
oftenbattery-poweredorhavelimitedaccesstopowersources. onthedesign,andthetargetuserdemographic[506].Crafting
Local computing tasks must be optimized to minimize energy interfacesthatarebothintuitiveandfunctionalacrossdifferent
consumption and extend device battery life [498]. Some IoT devices and user groups is a complex challenge [507], [508],
applications require real-time processing of data for timely [509], [510]. Moreover, ethical concerns about how data col-
decision-making. Ensuring that local computations meet the lectedbyIoTdevicesisusedcanarise,includingissuesrelated
requiredresponsetimeswithoutintroducinglatencyiscrucial. tosurveillanceordiscriminationbasedondataanalytics[511].
Therefore, the complexities of algorithms in local computing Ensuring ethical use while maintaining utility is a challenge
would be limited, therefore potentially restricting the comput- that has social, cultural, and political implications.
ing accuracy. Nevertheless, it is expected that with improved Addressing these challenges in IoT will require a multi-
algorithm optimization efficiency and more advanced power- disciplinary approach, involving contributions from various
supplying devices, the challenges in data storage and comput- fields including cybersecurity, legal studies, environmental
ing can be effectively mitigated or resolved. science, and social sciences. While these challenges present
In the realm of security and privacy, one of the pressing significant obstacles, they also open avenues for innovation
concerns is data encryption. As IoT devices proliferate, creat- and advancements in technology and policy.
ing secure methods to encrypt data across a network becomes
increasinglycomplex[499].Traditionalencryptiontechniques V. CONCLUSION
may not be applicable, given that many IoT devices have
Inthissurvey,wediscusshowArtificialGeneralIntelligence
resourceconstraints.Deviceauthenticationfurthercompounds
(AGI) can revolutionize the Internet-of-Things (IoT) systems
this issue. In a sprawling IoT network, establishing trust
and frameworks, and at the same time introduce unique
between various devices and central servers is complex and
challenges. AGI can be used to apply knowledge gained
critical. A single unauthenticated or compromised device can
in one domain in the heterogeneous data collected in IoT
becomeagatewayfornetwork-widevulnerabilities.Alongside
systems to new domains, integrate and contextualize data
this is the question of user privacy. IoT devices often collect
from diverse sensory inputs, understand causes and effects in
and transmit personal information, from health data to geolo-
events, discover, communicate, and interact with humans on
cation.Howthissensitivedataisstored,accessed,andutilized
dynamicandcomplexsystems,andlearnandapplynewskills
is a subject that demands careful consideration, both from a
to unseen tasks. AGI-powered IoT systems have a wide range
technical and ethical standpoint [500].
of applications including smart grids, home, agriculture, man-
Scalability and interoperability pose another set of chal-
ufacturing, healthcare, and transportation. In the meantime,
lenges [501]. With IoT devices coming from a plethora of
those applications require significant new research in AGI to
manufacturers, each with its own set of standards and pro-
be adapted to resource-limited scenarios in IoT. Data privacy
tocols, achieving seamless interoperability is a daunting task.
and security also need to be considered in designing AGI-
This lack of standardization is not merely a technical obstacle
powered IoT systems.
but a barrier to scaling IoT networks to accommodate the
While this article summarizes the diverse opportunities and
exponential increase in devices. In the absence of universal
applications enabled by AGI, we expect to see new and
protocols, interoperability, and scalable architectures become
exciting applications powered by AGI in IoT with the rapid
significant hurdles, inhibiting the efficient growth and func-
evolution of technology in AI and machine learning. We hope
tionality of IoT ecosystems.
this survey can inspire researchers from both the AI as well
Environmental and social challenges are often overlooked
as the IoT communities to develop and explore the limitless
but are no less significant [502]. While IoT devices offer
possibilities of AGI+IoT.
potential solutions for environmental monitoring, they them-
selves can be quite energy-intensive, thus contributing to the
very problems they aim to solve. This irony underscores the
REFERENCES
need for designing more energy-efficient devices. The rapid [1] K. Lakhwani, H. K. Gianey, J. K. Wireko, and K. K. Hiran, Internet
development and turnover of IoT technology also lead to of Things (IoT): Principles, paradigms and applications of IoT. Bpb
Publications,2020.
another concern: electronic waste [503]. The limited lifespan
[2] J.H.Nord,A.Koohang,andJ.Paliszkiewicz,“Theinternetofthings:
of these devices exacerbates existing environmental issues Reviewandtheoreticalframework,”ExpertSystemswithApplications,
related to e-waste, a challenge yet to be fully tackled. vol.133,pp.97–108,2019.
[3] A. A. Laghari, K. Wu, R. A. Laghari, M. Ali, and A. A. Khan,
Regulation and governance in IoT are largely nascent and
“A review and state of art of internet of things (iot),” Archives of
nebulous[504].Despitetherapidtechnologicaladvancements, ComputationalMethodsinEngineering,pp.1–19,2021.
the legal framework governing IoT remains ambiguous, creat- [4] L. Farhan, S. T. Shukur, A. E. Alissa, M. Alrweg, U. Raza, and
R.Kharel,“Asurveyonthechallengesandopportunitiesoftheinternet
ing challenges for accountability and compliance [505]. This
ofthings(iot),”in2017EleventhInternationalConferenceonSensing
isparticularlycriticalwhenIoTtechnologiesareimplemented Technology(ICST),pp.1–5,IEEE,2017.
in sectors with stringent regulations, such as healthcare [494] [5] R.LohiyaandA.Thakkar,“Applicationdomains,evaluationdatasets,
andresearchchallengesofiot:Asystematicreview,”IEEEInternetof
and transportation. The absence of clear regulatory guidance
ThingsJournal,vol.8,no.11,pp.8774–8798,2020.
canleadtoethicalandlegalgreyareas,puttingbothusersand [6] L.Atzori,A.Iera,andG.Morabito,“Theinternetofthings:Asurvey,”
providers in precarious positions. Computernetworks,vol.54,no.15,pp.2787–2805,2010.
[7] K.-H. N. Bui, J. J. Jung, and D. Camacho, “Consensual negotiation-
Last but not least, human factors should not be underesti-
baseddecisionmakingforconnectedappliancesinsmarthomeman-
mated.TheusabilityofIoTdevicescanvarywidelydepending agementsystems,”Sensors,vol.18,no.7,p.2206,2018.