IoT platforms assessment methodology for COVID-19 vaccine logistics and transportation: a multi-methods decision making model

The supply chain management (SCM) of COVID-19 vaccine is the most daunting task for logistics and supply managers due to temperature sensitivity and complex logistics process. Therefore, several technologies have been applied but the complexity of COVID-19 vaccine makes the Internet of Things (IoT) a strong use case due to its multiple features support like excursion notification, data sharing, connectivity management, secure shipping, real-time tracking and monitoring etc. All these features can only feasible through choosing and deploying the right IoT platform. However, selection of right IoT platform is also a major concern due to lack of experience and technical knowledge of supply chain managers and diversified landscape of IoT platforms. Therefore, we introduce a decision making model for evaluation and decision making of IoT platforms that fits for logistics and transportation (L&T) process of COVID-19 vaccine. This study initially identifies the major challenges addressed during the SCM of COVID-19 vaccine and then provides reasonable solution by presenting the assessment model for selection of rational IoT platform. The proposed model applies hybrid Multi Criteria Decision Making (MCDM) approach for evaluation. It also adopts Estimation-Talk-Estimation (ETE) approach for response collection during the survey. As, this is first kind of model so the proposed model is validated and tested by conducting a survey with experts. The results of the proposed decision making model are also verified by Simple Additive Weighting (SAW) technique which indicates higher results accuracy and reliability of the proposed model. Similarly, the proposed model yields the best possible results and it can be judged by the precision, accuracy and recall values i.e. 93%, 93% and 94% respectively. The survey-based testing also suggests that this model can be adopted in practical scenarios to deal with complexities which may arise during the decision making of IoT platform for COVID-19 SCM process.


Motivation
• The major motivation of presenting the proposed evaluation framework for logistics and transportation of COVID-19 vaccine is that we failed to find any work in well-reputed journals of supply chain management and logistics domain that is focusing on the evaluation of IoT technologies towards the L&T.However, the work presented by Orji et al. 20 is targeting only on the prioritization of factors (features) related to the deployment of Blockchain technology in the freight logistics industries.They have applied only ANP to rank the features for adoption of Blockchain technology.In comparison to our proposed decision making methodology, the proposed model supports multi-methods and applies hybrid MCDM approach towards the evaluation and selection of IoT platform for L&T of COVID-19 vaccine.• The L&T companies are facing exorbitant challenges to deliver the COVID-19 vaccine to the target location but the existing technological solutions are not sufficient to provide meaningful solutions.This evaluation framework will provide them a best option to adopt the right IOT based solution for the industrial use and use the optimal features by choosing the best IoT platform.• We failed to identify any work that is focusing on the IoT platform evaluation and features taxonomy.Simi- larly, the supply chain managers have less technological background and they are unaware about the IoT platform market and evolving features.Therefore, we were motivated to present this work to provide solutions towards selection issues in IoT platform for shipping of COVID-19 vaccine.
This paper is composed of eight (8) remaining sections: "Related work" Section is related to the literature study."COVID-19 vaccine and supply chain managment process" Section discusses the COVID-19 vaccine SCM process cycle in detail."IoT platform evaluation framework for COVID-19 SCM" Section highlights different kinds of issues faced by SCM in its every phase and presents IoT evaluation and decision making model."Framework evaluation and testing" Section highlights the testing and verification of proposed decision making evaluation model."Proposed work validation" Section is about the validating the proposed model."Managerial implications" Section is related to managerial implications of this work."Study limitations" Section explains the limitations of this research "Conclusion" Section ends with conclusion of this research work.

Related work
In our literature study, we failed to identify those studies that are focusing on identifying issues related to COVID-19 vaccine transportation and logistics deliver and providing solutions based on using IoT platforms.However, in existing literature different analysis models have been proposed to perform analysis of COVID-19 related data [21][22][23] .Some evaluation models utilizes statistical approaches to deal with COVID-19 related data 24,25 while some are AI-based algorithms using Artificial Neural Networks (ANNs) to build simulation model and produces the promising results related to COVID-19 analysis 26 .However, he main focus of this research is to highlight those research studies that are focusing on the analysis and issues related to COVID-19 vaccine during the shipping process.This research collects those studies that are intended to perform the analysis and evaluation of IoT platforms in SCM domain.Literature study of this research work falls into two different categories.In first category, we have identified the different issues and challenges in L&T and in second category, we shed light on decision making approaches related to the selection of IoT platforms in SCM area.
Younan et al. 27 addressed the challenges and proposed recommended technologies such as ICT in IoT layers in an industrial environment.Ding et al. 28 highlighted the impact of IoT on supply chain management.This work unfolds the challenges in smart logistics based on IoT and provides research directions for the development of smart logistics.Thibaud et al. 29 studied IoT-based applications in the environmental, health and safety (EHS) industries.They highlighted challenges in IoT adoption in high-risk EHS industries and proposed solutions to these challenges.The literature review presented by Golpîra 30 identifies the research gaps and provides guidelines for future research based on the presenting architecture known as Logistics Internet-of-Things (L-IoT).Similarly, Manavalan et al. 31 presented a conceptual framework based on five management parameters: business, management strategy, sustainable development, technology and collaboration.They explored the prospective opportunities available in IoT-embedded sustainable logistics for Industry 4.0 transformation.Wang et al. 32 presented IoT-based intelligent logistics system based on cloud and robotics technologies.They applied ant colony and Dijkstra algorithms for supply chain operations.
We also focused on to report the decision making approaches with respect to selection of IoT platform in different sectors to meet the business needs.In this regard, the Kondratenko et al. 33 evaluated seven (7) IoT platforms by using MCDM approach and among these platforms, they ranked Kaa platform as a best choice for IoT applications.They used eight (8) criteria for decision making purposes such as reliability, device management, level of integration, database functionality, data analytics, protocol support, processing level and action management and visualization usefulness.Similarly, the work presented by M. Ullah et al. 34 is also based on features and platforms evaluation.They evaluated five IoT platforms such as AWS, Microsoft Azure, IBM Watson, Google cloud an Oracle IoT.They formulated criteria based on twenty-one key features and after the assessment using Delphi method for criteria categorization they concluded that among the selected platform the AWS platform is best of the all.Lin et al. 35 applied AHP and probabilistic linguistic approach by taking five (5) IoT platforms into account based on features such as scalability, security, usability, market longevity, integration flexibility, pricing and availability.Contreras-Masse et al. 36 evaluated three IoT platforms based on varying features by using AHP and PROMETHEE-II approach by selecting the Azure IoT as a more suitable and rational choice for IoT solution.Orji et al. 20 evaluated the factors method that influence the adoption of Blockchian technology by using Analytic Network Process (ANP) method.Mijuskovic et al. 37 applied AHP method for evaluating five (5) IoT platforms with respect to many functional requirement.Similar works presented by authors for are 38,39 .

COVID-19 vaccine and supply chain managment process
The COVID-19 vaccine SCM process is a complex process as it is not a single activity but composed of a series of different steps/procedures to deliver the vaccine from the manufacturing spot to the final delivery location.A complete picture of the different phases and entities involved in the COVID-19 vaccine SCM model is shown in Fig. 1.
Following are the major actors involved in COVID-19 SCM cycle.
• Supplier The COVID-19 supply chain process starts with suppliers, who deliver raw materials to the manu- facturers.• Manufacturers They receive orders from wholesalers or distributors, and then the required amount of vac- cines is delivered to them after the completion of production.• Wholesalers or distributors These are the people who deliver the product to hospitals or pharmacies • Hospitals or pharmacies These are the stakeholders who buy products from wholesale dealers.
• Patients This is the last point, where vaccines are administrated to the patients.

IoT platform evaluation framework for COVID-19 SCM
The number of IoT platforms has significantly elevated in the recent times.This rapid rise has led to the selection and decision making issue in area of supply chain management.In the current pandemic situation, the selection of right IoT platform that can be used as technological solution of organizational needs is tough task due to bulky list of available platforms employed in L&T system.The selection of right IoT platform also has major impacts on COVID-19 vaccine shipping as it requires technical skills and expertise to choose the right choice among the list of available IoT platforms.The selection and decision making related to IoT platform for supply chain management of COVID-19 vaccine should not be done haphazardly but it should be done based on quantitative data and empirical assessment procedures.To address these selection issues and problems, we propose evaluation framework also known as decision making model for selection of the best IoT platform in L&T of COVID-19 vaccine.The step wise detail of building the proposed evaluation framework is depicted in Fig. 2. Following are the major phases of the proposed evaluation framework of IoT platforms for COVID-19 vaccine L&T.

Identifying issues in COVID-19 SCM and building criteria
Issues are identified based on literature study and highlighted during the SCM of corona vaccine.The entire SCM process of COVID-19 is hampered by various hiccups and bottlenecks.Different issues need to be addressed for the smooth running of the supply chain process.Unfortunately, these challenges are involved in every phases of SCM such as manufacturing, distribution and transportation, CCM and administration of the COVID-19 vaccine.

Manufacturing issues
Vaccine development plays a pivotal role in combating infectious diseases 40 .However, the vaccines are the most complex products as they require hundreds of components for manufacturing.The development of coronavirus vaccines and auxiliary supplies have resulted in a formidable universal challenge.There are a number of pharmaceutical and manufacturing companies that are racing to develop the most effective vaccines.However, vaccine quality should be a major focus of vaccine manufacturing companies.The process of developing vaccine is impeded by many challenges.It is important to identify these challenges and provide tentative solutions to increase the production rate.During the literature study we identified many challenges but the quality is major concern, when it comes to manufacturing of COVID-19 vaccine.The manufacturer of the vaccine must ensure that each vaccine is based on consistent quality and requires repeated testing procedures.Vaccine production is a biological process; therefore, some batches of vaccine may fail for reasons that are not always clear and ultimately it may lead to further delay in production 41 .There are quite few manufacturers in the world that can produce vaccines on a large enough scale to meet the needs of a pandemic.Pfizer, Moderna, and other pharmaceutical companies comprehensively addressed the challenges related to the manufacturing of vaccines faced by researchers and pharmaceutical experts.

Logistics and transportation issues
The shipment of the COVID-19 vaccines from the production point to the final retailers over a few days is a complex system due to many factors like storage factories, airplanes, cargo stations, warehouses and so on 42 .To avoid more waves of the COVID-19 pandemic, it is important to address all the issues related to the logistic, transportation and distribution of vaccines.According to a report by the logistic firm Boyle Transportation, it is estimated that 14 billion COVID-19 doses will be required for two doses per person on earth, which is itself a challenging task 43 .Considering the deployment of the COVID-19 vaccine, a transparent distribution and implementation are indispensable for vaccines.Successful distribution and deployment will result from the collaboration of complex networks of governing agencies, healthcare workers, the public and companies 44 .

Cold chain management (CCM) issues
During the current COVID-19 pandemic, the main challenge was to keep the integrity of vaccine.COVID-19 vaccine are temperature sensitive and they require ultra-care to be properly monitored during its shipping.The integrity of the vaccine is solely dependent on cold chain management.The most important part of any COVID-19 vaccine program is the cold chain.Infect, the success of the SCM process hinges on the correct implementation of the CCM.The efficacy of the vaccine is achieved by maintaining the cold chain process in the right way.The cold chain unit ensures that cold chain storage facilities are provided from the manufacturing to the administration site of the vaccine.According to a report released by the WHO, it is estimated that 2.8 million doses of vaccines were lost in 2011 in five countries due to the failure of the cold chain.Cold chain solitary contributes to 80% of all vaccination costs 45 .The most common temperature for storing vaccines ranges from 2 to 8 degrees but some of the coronavirus vaccines may require storage temperatures up to − 70° centigrade.Storing vaccines at up to low temperatures is a big challenge for cold chain storage infrastructure and it is ought to be addressed to keep the vaccine under thermal requirements.

Issues in vaccine administration
While administrating the corona vaccine, still there are some challenges that are required to be addressed to complete the entire cycle of SCM.These major issues include like lack of awareness, people reluctance to vaccine, lack of ancillary supplies etc.The complete details of challenges and bottlenecks during each phase of the COVID-19 vaccine SCM process are given in Table 1 41,[46][47][48][49][50][51] .
A full-pledged technological solution to handle the problems of vaccine logistics and transportation is indispensable.After identifying issues in the SCM phases, it is important to provide reasonable solutions towards these challenges and issues.The major focus of this research work is to provide IoT-based solutions by bringing the most viable IoT platform (discussed in next section).
After selecting alternatives, the next step of the proposed framework is to set the benchmark or criteria for selection of IoT platforms.In this step, we also selected the most famous and well-known IoT platforms for evaluation and ranking purpose.The criteria are designed based on the features/parameters of IoT platform.Therefore, we performed features-analysis to include the most relevant features that are covering all IoT business needs.The parameters selected from literature and they are consulted with the expert panel in IoT.The detail about the the finally selected parameters with the literature source is given in Table 2.
These parameters are selected based on the frequency of occurrence in the existing literature.The complete detail about the number of times a particular feature of IoT platform used for evaluating the IoT platform is given in Fig. 3.
For any platform, it is important to have certain features and any certain features is required to be supported by every possible IoT platform.Every platform is interdependent on every feature and vice versa.The hierarchical structure of "n" number of IoT platforms with "n" number of features are given in Fig. 4.
After finalizing IoT platforms alternatives we conducted a survey for collecting data.In this regard, a case study has been conducted for collecting data related to selected platforms and verifying the criteria features.In this cases study, the IoT platform vendors along with SCM experts were consulted in finalizing the list of important features.This case study followed a well-known method such as Estimate-Talk-Estimate technique which is also known as Delphi method.This technique is applied to check the criteria features by consulting the During the survey the experts' opinions were recorded about the selected platforms against the designed criteria.The expert groups were asked to provide response based on numeric values.They provided response about every platform with respect to criteria features based on Saaty's scale which is ranging from 1 to 10.All the expert groups/participants provided informed consent for participation in this study.All the methods were carried out by following the relevant guidelines and regulations.The outcome of the final activity after using ETE techniques is a tabular data consisted of IoT platform and features.This data will be provided as input in the decision matrix form in next step.Data related to decision matrix about the selected IoT platforms against the criteria feature is given in Table 3.The next step was to assign weights to criteria features using CRITIC and evaluate and prioritize IoT platform by using TOPSIS.

Assigning weights to feature/parameters
The values assigned by expert panel during the case study may be suffering from biasness and subjectivity, therefore our focus is to avoid the element of biasness in our judgment.For this purpose, we applied the most famous MCDM method known as CRITIC.This method uses mathematical procedure and statistical approaches to solve any MCDM problem.The first step of CRITC method is to build a decision matrix.The data about the decision matrix (D ij ) is derived from Table 4 as given below.The input provided to CRITIC method is a decision matrix, where mathematical calculations are performed to obtain weights to the features as shown in Table 4.
The output obtained from the CRITIC method indicates that the highest weightage is assigned to scalability feature, followed by cost, deployment and protocol with equal numeric value of weights.The complete detail of weights assigned to the features is visually depicted in Fig. 6.

IoT platforms evaluation and prioritization
The CRITIC method has been applied for assigning weights to the features of criteria.After giving weights to the features of IoT platform, next step is to evaluate and rank IoT platform based on the designed criteria features.TOPSIS method has been applied to evaluate the platform alternatives based on criteria.The stepwise detail of TOPSIS method is given below as.

Making decision matrix
In first step of TOPSIS method the decision matrix is formed with "n" number of criteria and alternatives.The decision matrix denoted by "D" is built by using Eq.(1).Decision matrix is normally built with the assistance of expert groups.

Building normalized decision matrix
The data given in decision matrix (D) originates from different sources, therefore, it has to be normalized by converting into a dimensionless matrix.The comparison of different criteria is done via dimension matrix.A normalized decision matrix (R ij ) is built by using the following Eq.( 2).To negate the element of the biasness the decision matrix is normalized.
Step-3.Determining the weighted normalized decision matrix.Sometimes, all attributes may not be same values or importance.In this regard, "V" is calculated which indicates the values of weighted normalized decision matrix.It can be computed by the multiplication of every element (R ij ) of normized decision matrix with a random weight number as follows using Eq.(3).
Step-4.Finding ideal positive and negative solutions.The positive ideal solutions represented by A + and negative eal solutions denoted by A − are determined by using previously calculated weighted decision matrix.They are found by using the following Eqs.( 4) and ( 5) respectively.
where "J" denotes the beneficial attributes and "J' " is shows non-beneficial attributes.For complete understanding ideal positive (A + ) and ideal negative point (A − ), the visual representation is given in Fig. 7 76 .
Step-5.Determining the separation measures.Ideal separation measure is denoted by S + and no-ideal separation measure is represented by S − which are obtained with the help of using the following two mathematical equations.
(2)  Step-6.Finding of relative closeness.The relative closeness symbolized by C i variable is impoant for the purpose of final ranking and it is found with the help of the Eq. ( 8) as given below.
Step-7.Ranking of alternatives.The alternatives were ranked by using the value of C i i.e. the alternative with higher value of C i value stands higher in ranking and performance and alternative having low value of C i is considered as low in performance and significance.Ranking of alternatives or preferences can be done in both descending and ascending order.After assigning weights to the features, the TOPSIS approach is applied to rank the IoT platforms based on the criteria by performing mathematical and statistical calculation.The application of TOPSIS produces numeric results that can be used for ranking of IoT platforms.TOPSIS method uses a sequential procedure for ranking the IoT platform alternatives.It ranks the IoT alternatives based on the values of performance index or relative closeness (C i ).The higher value of C i indicates that alternative may be considered as best choice among the comparing list of alternatives.The decision matrix already obtained is provided as input to the TOPSIS.This method uses Eq. ( 3) for converting decision matrix to normalized form.The major purpose of this procedure is to convert the data into a form that is free from personal biasness.Then, Eqs. ( 4) and ( 5) have been applied to obtain ideal positive and ideal negative solutions.The detail of all IoT platform alternatives against the features with respect to ideal positive and ideal negative solutions is given in Fig. 8.
All the values of every platform against the criteria are between ideal positive and ideal negative solutions.Hence, our results are accurate and we can proceed to the next step.The separation measures (S + , S − ) and relative closeness (C i ) are calculated by using Eqs.( 6), ( 7) and ( 8) respectively and output is given in Table 5.The value of C i is i5mportant consideration as it decides the outcome of the evaluation process i.e. the higher value of C i indicates the best solution and lower value shows the worst cases.According to the results of Table 5, P3 alternative has the highest value of all the IoT platforms.
So, Microsoft Azure (P3) is reckoned as best choice among the list of IoT platforms based on our proposed criteria.The comparison of all IoT platforms selected in this study is given in Fig. 9.
Thus, the finding of this study suggests that Microsoft Azure IoT platform is the right candidate for shipping of COVID-19 vaccine.It has the ability to support multiple features in respect of supply chain operation and logistics process.This platform uses different technologies such as Azure cloud, Azure IoT hub, Azure machine learning and Azure maps to provide smart delivery in L&T.Microsoft Azure IoT provides smart transportation infrastructure, assessing the road conditions, real time and historic traffic management.It uses real time data to bring fleet operation or smart logistics by sending alerts, monitoring performance, optimizing the delivery routes ( 6)

Framework evaluation and testing
It is important to evaluate and test the IoT evaluation framework of IoT platforms.The proposed evaluation framework is checked by the experts for the features included and excluded during the criteria designing.To achieve this, we conducted a case study, in which expert participated and they presented their opinions about the number of features and gravity of features of IoT platforms selected for evaluation.We evaluated framework by two different approaches such as evaluation by experts i.e. checking the criteria feature and survey based evaluation which is focused on assessing the overall performance of proposed framework.The detail of both assessing methods is given as.

Evaluation by experts
After finalizing and building the evaluation framework, we evaluated the features of the proposed framework by consulting the experts group due to the theoretical nature of framework.This framework is evaluated by three evaluation parameters such as accuracy, precision and recall.As, said earlier that the features are the building block of IoT platform.Therefore, it is important to check relevant, irrelevant, recommend and not recommend features to keep the framework working well and producing the desired outcomes.For evaluation purpose, we used four type of variables to classify the features such as the number of features recommend by our experts group and the proposed framework is denoted by "a".The number features only suggested only by proposed framework are represented by variable "b".The number of features only suggested by expert panel is shown by "c".Similarly, the variable "d" indicates the number of features not suggested by evaluation framework and nor suggested by expert group.This method used for evaluation and testing of evaluation framework is well-known technique often used to evaluate the context-base recommendations systems 78,79 .The evaluation parameters such as accuracy, precision and recall are obtained by using the following Eqs.( 9), ( 10) and ( 11) respectively.
The feature classification in terms of recommended, not-recommended, relevant and irrelevant is given in Table 6.
The complete process of finding the evaluation parameters expert panel in comparison to our evaluation framework is given in Table 7.

Survey based evaluation
Our survey based evaluation in this study is inspired from the study conducted in 78 .In this step to evaluate the IoT platform evaluation framework is through conducting a survey.In this survey, three expert group participated who are currently enrolled in Phd and MS degree programs.The framework evaluation through survey is conducted by using a five-points scale.In this survey, we asked 28 research questions of different categories and  8.We tested the proposed IoT evaluation framework based on a questionnaire consisted of many questions related to evaluation parameters such as security, usability, effectiveness and information and knowledge.The complete detail of questions/parameters asked during the survey and response of each group in terms of numeric values are given in Table 9.
After collecting responses from the expert's groups in Table 8, it has become quite clear that mean values of all groups are more than 4. It means that all the expert groups are satisfied with our proposed IoT decision making model.All the group members have shown positive feedback and are in favour of recommending this evaluation framework for choosing the right IoT platform selection for L&T of COVID-19 vaccine.

Proposed work validation
The survey based testing and verification were further validated by using a well-known MCDM techniques known as SAW method.This method has been applied to check the consistency and accuracy of our proposed decision making model.This method is very simple method.This method works based on the following major steps 80 .
The following equation is used to achieve the normalization of data.
• Step-2.Weight assignment to each criterion  As we have already computed the criteria weights by using CRITIC method so we skipped the first two steps in this technique.Equation ( 14) is used to calculate the ranking score for each IoT platform alternative.The results calculated through SAW approach in comparison to our proposed methodology are given in Table 10.
The results obtained after the application of SAW approach are also indicating that P 3 platform has the higher value of ranking score in the list of selected IoT platforms.So, it is concluded from this empirical data that quantitative results are accurate and consistent.Hence, the proposed decision making model for IoT platform is also validated.The comparison of all selected IoT platforms based on the ranking score calculated by using SAW approach is given in Fig. 11.According to SAW approach, Microsoft Azure IoT platform stands on the top of all selected platform.( 14)

Managerial implications
In this section, we are highlighting the recommendation and implications of the impacts of the proposed decision making model and selection of right IoT platform in logistics and transportation of COVID-19 vaccine from the perspective of managers and IoT platform developers.The proposed decision making model will enable the supply chain, cold chain, fleet, transportation and logistics managers to adopt the important aspects of IoT adoption or technology as whole or in incremental shape.As, the selection of IoT platform has influential impacts on the business operations of supply chain managers.Managers can adopt IoT opportunities to address bottlenecks in the conventional L&T vaccine delivery.The right selection of IoT platform helps managers to shift the product safer and efficient.IoT solutions help managers track the real-time visibility and traceability of vaccines by using various IoT sensors.These sensors are placed on pallets and have the ability to read light, temperature, humidity and other manufacturing details of the vaccine by sending alerts from the manufacturing point until the last mile logistics delivery.According to finding of this study, the selected IoT platform i.e.Microsoft Azure IoT provides features like reliability, efficiency, location intelligence, enhanced safety and low cost delivery through smarter ways.
The results of the presented decision making model provide a better and deeper understanding about the features selection and their importance.This study introduces the most important features about IoT technology adoption in L&T and helps in evaluating features based on their relative importance.The evaluation criteria feature defined in this research work is the most compact and cover all the dimensions of IoT platforms.Thus, this framework will help the L&T managers not to get worried about the features selection of IoT platform.For example, the device management feature of IoT platform allows the supply managers to take in to account the devices, sensors and actuators before applying them as business solution.According to the finding of our study the selected IoT platform such as Microsoft Azure IoT platform allows the IoT sensors management that provides geolocation services which means the cargo or vehicles or other assets can be monitored over a wide geographical distance in harsh environments.The selected IoT platform by this model also provides fleet management such as vehicle and driver and rout information.
The proposed decision making model will enable the fleet and supply managers to include the most critical features related to cold chain management as the COVID-19 vaccines are temperature sensitive.Therefore, the evaluation and decision making of IoT platform is imperative due to the significance and complexity of cold chain management.The deployment of right IoT platform will enable the supply chain managers to keep the track of temperature of COVID-19 vaccine in the route and to keep vital information about the real time status of vaccine shipping.
The supply chain managers are not the technical people in terms of computer literacy.Therefore, the selection of right IoT platform vendor requires more technical and expertise in this domain.Similarly, the number of IoT platform are drastically changing and the emergence of these platforms become a major issue for SCM managers to go through the extensive decision making process and selection the more suitable IoT platform.This evaluation framework is helpful not only for supply chain managers to deploy the best IoT platform for their business requirements but will also help the platform developers to develop, customize and integrate the most important features especially related to the shipping and logistics.It will also help the developers to incorporate all the features and functionalities based on their relative importance in IoT platforms.This decision making model platform will help them to update the IoT platform or customize it to provide up-to date services and features related to the L&T system.
In nutshell, the proposed decision making model will enable all the stakeholders working in logistics, transportation, supply chain, fleet; government agencies and platform developers to bring the advancement in the successful adoption of IoT technology for COVID-19 vaccine delivery to meet the emerging requirements of market.

Study limitations
The proposed methodology produces promising results, covering all features and enhancing the decision making process but still are some limitations that may affect the practicability and stability of the proposed model as given.
• The criteria elements or parameters selected in this study are not absolute but are relative.It means that it is not mandatory that other models will also follow the selected features in their evaluation methodology.However, these are the best possible features supported by IoT platform with respect to deploying IoT technology for COVID-19 vaccine logistics and shipping.• This methodology is consisted of TOPSIS, CRITIC and SAW approaches for evaluation purpose but still we believe that the proposed model can enhanced in by using the concept of fuzzy algorithms.• The proposed model is the first time presented so authors feel that as more models will introduce in this domain then the proposed model can be better compared with the results.This is the main reason that we validated the proposed model using two different approaches such as survey-based evaluation and using SAW techniques to verify the quantitative results yielded by this model.

Conclusion
The COVID-19 vaccine delivery from the point of manufacturing till the last mile delivery can be challenging job due to many factors involved such as cold storage management, safe delivery, quick delivery, temperature monitoring, improper routing etc. Different technological solutions have been suggested but the role of IoT is dominant.The success and operation of IoT hinges around the deployment of right IoT platform.Therefore, many IoT platforms have been introduced in the market.This proliferation resulted in decision making problems for SCM managers, who have less knowledge or expertise about the technology deployment and selection of right technology to perform various tasks easily and efficiently.As, there is no mechanism to compare directly the features of various IoT platforms due to the IoT platforms evolution and number of features supported.The first step of the proposed model is to define criteria of evaluation for selected twenty (20) IoT platforms.Data collection process about the IoT platforms was supported by using ETE or Delphi technique.This model adopts the hybrid decision making approach and performs the quite accurate empirical analysis of IoT platforms by using CRITIC and TOPSIS.The proposed methodology prioritizes IoT platforms according to defined criteria.This model ranks Azure IoT platform at the top for SCM of COVID-19 vaccines.This platform can be installed by the SCM management team to perform their routine tasks related to vaccines management.Thus, the proposed model assists in selecting the most suitable and ideal option of IoT platform among the IoT offerings for the logistics and supply chaining of COVID-19 vaccines.This IoT platform can be leveraged to provide excursion notifications, vaccines monitoring, cold chain management, location awareness, routing tracking, shipping information etc.The first five model suggested by the proposed model are the right candidates for handling the The evaluation metrics such as accuracy, precision and recall for the features validation are calculated.The values of precision, accuracy and recall 93%, 93% and 94% respectively.It indicates that the features selected in the evaluation criteria are covering the most essential aspects of IoT platforms.The overall performance of the proposed model is tested and checked based on five parameters such as security, usability, information and knowledge and effectiveness.This is survey based testing and results of this procedure suggests that the proposed assessment methodology can be used for evaluation and prioritization of IoT offerings that are intended for COVID-19 vaccine logistics and transportation.Hence, we conclude that this model is consistent and results are accurate and it can be applied to address the challenges faced in the logistics and transportation of COVID-19 vaccine.The model has the potential to perform well in practical scenario by tanking input of data about IoT platforms and it will evaluate platform based on the best possible features.It can save time and energy of managers by picking the most ideal IoT solution for their business needs.This model has the calibre to provide guideline for the IoT platform developers to update the functions and features of IoT platform with respect to vaccine management.In future work, the authors are looking forward to apply the advanced and fuzzy decision making techniques for evaluating IoT platforms such as Fuzzy CRITIC, AHP and Fuzzy TOPSIS for designing a more sophisticated decision making model.Similarly, the evaluation parameters can also be updated to add more or new features of IoT platforms which are rapidly evolving due to the emerging situations in logistics and transportation of COVID-19 vaccine.
https://doi.org/10.1038/s41598-023-44966-yexperts group in the field of IoT.They provided their valuable response about the features selected in this study.The responses or opinions from the expert's panel are collected based on this technique.Intermediate results are obtained after applying ETE technique in first round.Second round completes the collection of data about the IoT platforms with respect to criteria.Our focus is to collect the most relevant data based on the issues (already discussed) in selection of IoT platforms.The complete detail of applying this technique in context of receiving responses from expert's panel is given in Fig.5.

Figure 4 .
Figure 4. Hierarchy and inter-dependencies of features and IoT platforms.

Figure 7 .
Figure 7. Ideal positive and ideal negative points.

Figure 8 .
Figure 8.Comparison of IoT platforms alternatives based on A + and A − .

Figure 10 .
Figure 10.Microsoft Azure IoT platform features of logistics and transportation.

Figure 11 .
Figure 11.IoT platform ranking score by SAW method.

Table 2 .
Evaluation parameters description and literature sources.

Table 7 .
Recommendation evaluation parameters results.

Table 9 .
Responses of expert's groups.

Table 10 .
Proposed work comparison with SAW technique.
Covid-19 vaccine.The results of proposed model are validated and verified in two different ways.The selected features in criteria are verified by using a survey with experts and the results are confirmed by SAW methods and expert groups.