List of Publications from Thesis Work

Journal

1.      Rauf, Affan, Muhammad Nawaz, and Junaid Haroon Siddiqui. "Effective State Encoding for Breadth-First Generation of Complex Structures." IEEE Transactions on Reliability 68, no. 3 (2019): 1154-1167.

Available at: https://ieeexplore.ieee.org/abstract/document/8733194

Conference

2.      Rauf, Affan, Muhammad Nawaz, and Junaid Haroon Siddiqui. "Efficient iterative deepening for bounded exhaustive generation of complex structures." In Proceedings of the 40th International Conference on Software Engineering: Companion Proceeedings, ACM, 2018.

Available at: https://dl.acm.org/citation.cfm?id=3195002

Abstract: 

For centuries buildings have been a source of shelter and comfort for human beings. In their modern, contemporary form, buildings have even enabled us to artificially condition various elements, including air and water, to meet our comfort needs. Air comfort revolves around removing heat from the ambient air, while water comfort entails storing water for long periods of time for on-demand usage. But due to their high cost, not all buildings have state-of-the-art solutions for either comfort.

Air comfort is a challenge in the developing world due to the widespread use of legacy heating and cooling devices. These devices commonly lack thermal control units, causing temperature levels to exceed the required comfort range regularly. Such temperature is harmful to the health of the occupants and requires more energy to maintain it. Therefore, the heating and cooling process has become the largest consumer of energy in the household. Modern heating, ventilating, and air conditioning (HVAC) systems tackle some of these challenges, but the high cost of such systems impedes their penetration in under-developed countries. Inverted HVAC, practically an inexpensive IoT system, solves this problem by employing a greedy control algorithm that turns the device on and off on the edge of the comfort range, potentially leaving a large gap for further improvement. This thesis exploits this gap and proposes a data-driven model for improved Inverted HVAC control. The final result is an efficient duty cycling of thermal devices that maintains the optimum comfort level, avoids short cycling, and saves energy.

Like air, water is an essential ingredient of human comfort. However, the traditional water setups in homes expose the water tank to direct sunlight, which increases the water temperature to unbearable levels in summers. In ideal circumstances, water should be available at room temperature for various household activities. Although this problem has been tackled by modifying house designs in advanced countries, changing the design in developing countries like Pakistan requires a lot of investment beyond the financial means of the wide majority of the population. To ensure water delivery at room temperature, this work proposes pumping water from the ground tank to the roof tank at the time of usage. To enable this, a robust water usage forecasting model is essential. As part of our work, we develop and deploy custom IoT devices in homes to collect water usage data and develop a machine learning-based forecasting model. While grappling with the real-world data, we observed that the pattern of water, usage is subject to fluctuations attributed to guests and other one-off activities. To account for this dual pattern, we develop two models, one to forecast routine usage and the other for one-off outliers, allowing the system to dynamically interleave them based on the current usage pattern. Our custom IoT device is retrofitted into existing infrastructure with a hardware prototyping cost of $27, whereas it can save up to 16% on water heating costs, through a reduction in natural gas consumption, by leveraging ground tank’s water temperature while improving water comfort by up to 8 ∘C in summers and 3 ∘C in winters, on average.

Our proposed IoT systems can positively impact the lives of millions of people in developing countries.

Publications:

1. Samar Abbas, Ahmed Ehsan, Saad Ahmed, Sheraz Ali Khan, Tariq M. Jadoon, Muhammad Hamad Alizai: No-frillsWater Comfort for Developing Regions, ACM/IEEE International Conference on Information Processing in Sensor Networks (IPSN), Sydney, Australia, 2020. CORE RANK: A*

2. Samar Abbas, Abu Bakar, Yasra Noman, Khadija Hafeez, Ayesha Ali, Tariq M. Jadoon, Muhammad Hamad Alizai: Inverted HVAC: Greenifying Older Buildings, One Room at a Time, ACM Transactions on Sensor Networks (TOSN), 2018.

3. Samar Abbas, Ahmed Ehsan, Saad Ahmed, Sheraz Ali Khan, Tariq M. Jadoon, Muhammad Hamad Alizai: ASHRAY: Enhancing Water-usage Comfort in Developing Regions using Data-driven IoT Retrofits, ACM Transactions on Cyber-Physical Systems (under submission)

Abstract:

Progress on the UN Sustainable Development Goals (SDGs) is hampered by a persistent lack of data regarding key social, environmental, and economic indicators, particularly in developing countries. For example, data on poverty and slavery, the first and eighth of seventeen SDGs respectively, are both spatially sparse and infrequently collected in developing countries due to the high cost of surveys. Other such examples include taxation, population census, vaccine coverage and electricity demand forecasting. Use of satellite imagery provides an opportunity to automate these surveys, however, while considering large geographical area, several inherent complexities in satellite imagery make automated detection a challenging task. These include high amount of computation and storage as well as variation in images over time and space such as structural, environmental and sensor variations.

The main goal of this dissertation is to develop accurate and compute-efficient learning approaches for huge volumes of Earth's images. This study also addresses the problem of scalability of the model for automatic interpretation of satellite imagery covering a large geographical extent. For this purpose, we develop several deep learning strategies using spatial convolution, 3D convolution and graph convolution. In particular, we design four deep learning architectures: 1) Tiny-Inception-ResNet-v2, 2) Kiln-Net, 3) 3D Residual Network and 4) STAG-NN for spatial and spatio-temporal classification and detection for high resolution imagery. We also develop two largescale classification benchmark datasets from open-source remote sensing imagery of South Asia. The first spatial dataset, named Asia14, consisting of 14,000 Digital Globe RGB images and 14 classes including brick kilns, houses, roads, tennis courts, grass, farms, sparse forest, dense forest, orchards, parking lots, parks, oil tanks, and barren lands. The second spatio-temporal dataset, named 2C2D, consists of four key transition classes: construction, destruction, cultivation, and harvesting.

Land-use and Land-cover (LULC) is one of the essential steps in many remote sensing-based surveys. Initially, we developed a general purpose and compute efficient architecture for the problem of LULC classification and tested it on Asia14 dataset with 11 different classes. We purpose and employ Inception-ResNet inspired deep learning architecture called Tiny-Inception-ResNet-v2 which provide 63.8% reduction in parameters, and 2.6% increase in classification accuracy. We then used it for the localization of brick kilns to help eliminate poverty and bonded labour within the "Brick-Kiln-Belt" of South Asia. To incorporate intra-class variations over large-scale spatial analysis, we propose a coarse-to-fine strategy consisting of an inexpensive classifier and a detector that work in tandem to achieve high accuracy at a low computational cost. More specifically, we propose a two-stage gated neural network architecture called Kiln-Net. At the first stage, imagery is classified using the ResNet-152 model which filters out over 99% of irrelevant data. At the second stage, a YOLOv3-based object detector is applied to find the precise location of each brick kiln in the candidate regions.

To incorporate temporal variations over large-scale analysis, we then introduce a scalable method to predict socio-economic indicators using our spatio-temporal dataset 2C2D. Our approach improves existing techniques in three ways; forgoing hand-crafted features traditionally used in the remote sensing community, introducing a novel 3D spatio-temporal input technique and incorporating 3D convolutional kernels to explicitly model the spatio-temporal structure of the data and automatically learn useful features. We evaluate the efficacy of our approach on data from three different cities from different South Asian countries. Despite the significant computational overhead, 3D convolution networks fail to address the problems of non-Euclidean domains. We, therefore, develop Spatio-Temporal driven Attention Graph Neural Network (STAG-NN) to learn the interactions between dense spatial and sparse temporal data. Performance of the STAG-NN is evaluated for spatial and spatiotemporal feature learning over large scale analysis. We also demonstrate the comprehensive comparative analysis (in terms of accuracy and compute cost) of proposed techniques with state-of-the-art deep learning methods.

Our proposed solution will enable regional monitoring and evaluation mechanisms for the SDGs and will help governments target their interventions with the precision that was previously unheard of. Here we only demonstrate its application for the geo-localization of brick kilns and the detection of spatio-temporal transition classes e.g., construction, destruction, cultivation and harvesting. This research will also allow governments to monitor illegal kiln activities and implement strategies to help prevent famine and emancipate individuals trapped within institutions of slavery and support humanitarian efforts.

List of Publications:

• M. A. Bhimra, Usman Nazir, and Murtaza Taj. "Using 3d Residual Network For Spatio-Temporal Analysis Of Remote Sensing Data." IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2019. Link: https://ieeexplore.ieee.org/document/8682286 

• Usman Nazir, et al. "Tiny-Inception-ResNet-v2: Using Deep Learning For Eliminating Bonded Labors Of Brick Kilns In South Asia." IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops. 2019. Link: https://arxiv.org/abs/1907.05552 

• Usman Nazir, et al. "Kiln-net: A gated neural network for detection of brick kilns in South Asia." IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (JSTARS) 13 (2020): 3251-3262. Link: https://ieeexplore.ieee.org/document/9115879 

• Usman Nazir, Muhammad Awais Ather, and Murtaza Taj. "Detection of Illegal Kiln Activity During SMOG Period." International Conference on Robotics and Automation in Industry (ICRAI) 2023. Link: https://arxiv.org/abs/2301.07521 

• M Sohail, Usman Nazir, et al. “Analyzing the knock-on Impacts of 2022 Floods on Rabi 2023 Using Remote Sensing and Field Surveys” International Conference on Integrated Flood Management under Changing Climate Scenario (IFMCC), 2023. 

• Usman Nazir, et al. “Improved flood mapping using fusion of Sentinel-1, Sentinel-2 and Landsat-9 datasets” International Conference on Integrated Flood Management under Changing Climate Scenario (IFMCC), 2023. 

• Usman Nazir, et al. “Mitigating climate and health impact of small-scale kiln industry using multi-spectral classifier and deep learning” ICLR 2023 Workshop: Tackling climate change with AI.

• Usman Nazir, et al. “Spatio-Temporal Driven Attention Graph Neural Network with Block Adjacency Matrix (STAGNN-BA)” ACM Conference On Knowledge Discovery And Data Mining (SIGKDD), 2023. (submitted) 

Website: https://usmanweb.github.io/ 

Google Scholar: https://scholar.google.com.pk/citations?user=EKpcuAoAAAAJ&hl=en 

Abstract: 

The emerging cloud and edge computing infrastructure provides new opportunities to develop next- generation Internet-centered distributed applications that are adaptive, evolvable, and emergent. Such applications may include knowledge-driven distributed workflows that are dynamically orchestrated and managed by utilizing computation, data, and storage resources available in a cloud data center, enterprise networks as well as Internet of Things (IoT) devices. IoT devices in such workflows provide a diverse range of functionalities, from measurement and dissemination of sensory data observation to computation services for real-time data stream processing. In workflows that are designed for extreme situations, such as emergencies, a significant benefit of IoT devices is that they can help gain a more complete situational understanding of the environment. However, this requires the ability to effectively utilize resource constrained IoT devices.

The IoT-centric applications that we consider in this dissertation are knowledge-driven workflows. A unique aspect that differentiates them from traditional workflows (business processes, scientific workflows, etc.) is that they are emergent and their execution evolves based on the knowledge of the execution status, environmental context, and situation and case-specific parameters that are not known a priori and are subject to change at runtime. Moreover, in such workflows, the binding of tasks to service/resource endpoints may not be known at design time. The orchestration and management of such workflows therefore requires dynamic discovery, selection and binding of workflow tasks to available cloud or edge resources as well as establishing coordination between these resources based on the functional and nonfunctional requirements of the workflow.

In this dissertation, we develop an integrated framework that supports dynamic orchestration and management of IoT-centric and knowledge-driven workflow applications in the cloud and edge computing environment. Users submit their workflow orchestration and management requests to a workflow coordinator by providing workflow specifications. The different tasks in the workflow may run on computation/data resources on the cloud, edge nodes and on IoT devices. These edge nodes and IoT devices may be geographically distributed. For workflow orchestration, the coordinator dynamically binds the workflow tasks to the services available in the cloud or edge/IoT devices based on the location and context requirements of tasks. The binding may need to change in real time as the dynamics of the underlying environment change.

There are two main underlying approaches that perform dynamic orchestration and management in the proposed framework. The first approach is Global Orchestration and Management (GOM) approach, which is a centralized approach and it completely rely on Global Coordinator for binding, execution, deployment and adaptation of all the workflow related activities. The Global Coordinator maintains the global information of all the edge resources and cloud services. The second approach is Local Orchestration and Management (LOM) approach which, performs orchestration and management in a distributed manner by different peer nodes. A comprehensive experimental evaluation performed by emulating real-time data streaming workflows shows the effectiveness of our proposed approaches.

List of Publications:

1. Sehrish Amjad, Ahmed Akhtar, Muhammad Ali, Ayesha Afzal, Basit Shafiq, Jaideep Vaidya, Shafay Shamail, and Omer Rana, “Orchestration and Management of Adaptive IoT-Centric Distributed Applications”. IEEE Internet of Things Journal, vol. 11, no. 3, pp. 3779-3791, 1 Feb. 2024. doi:10.1109/JIOT.2023.3306238.

2. Sehrish Amjad, Ahmed Akhtar, Muhammad Ali, Basit Shafiq, Shafay Shamail, Ayesha Afzal, and Jaideep Vaidya, “Demo: Orchflow: Orchestration and Management of IoT Centric Distributed Workflows,” accepted for publication in the Proceedings of the 44th IEEE International Conference on Distributed Computing Systems (ICDCS), 23 July – 26 July 2024, Jersey City, New Jersey, USA.

Abstract: 

The use of blockchain technology has been proposed to provide auditable access control for individual resources. Unlike the case where all resources are owned by a single organization, this work focuses on distributed applications such as business processes and distributed workflows. These applications are often composed of multiple resources/services that are subject to the security and access control policies of different organizational domains. Here, blockchains provide an attractive decentralized solution to provide auditability. However, the underlying access control policies may be overlapping in terms of the component conditions/rules as well as events. Existing work cannot handle event-driven constraints and does not sufficiently account for overlaps in the policy leading to significant overhead in terms of cost and computation time for evaluating authorizations over the blockchain. In this dissertation, we develop an integrated framework for access control management of distributed business processes over blockchain. This framework allows generation of a composite access control policy for a given distributed business process based on the local policies of component services and reduces the policy evaluation cost over the blockchain. The local access control policies of component services may include both attribute-based and event-driven policies. The composite access control policy includes one or more smart contracts, which can be deployed on the blockchain for access control enforcement. The proposed framework supports composition and management of both attribute-based as well as event-driven access control policies. For attribute-based policies, we formulate a constraint optimization problem to generate an optimal composite access control policy. For event driven policies, we have developed an automata-theoretic approach that generates a cost-efficient composite access control policy. We reduce this composite policy generation problem to the standard weighted set cover problem, which is an NP-complete problem for which several approximation techniques exist. We show that the composite policy correctly captures all the local access control policies and reduces the policy evaluation cost over the blockchain. We have also proposed a game-theoretic approach for auditing access control enforcement in an efficient and cost-effective manner while incentivizing honest behavior of all parties. We have implemented the initial prototype of our approach using Ethereum as the underlying blockchain and empirically validated the effectiveness and efficiency of our approach. We have also conducted ablation studies to determine the impact of changes in individual service policies on the overall cost.

List of Publications:

1. Ahmed Akhtar, Masoud Barati, Basit Shafiq, Omer Rana, Ayesha Afzal, Jaideep Vaidya, and Shafay Shamail. “Blockchain Based Auditable Access Control For Business Processes With Event Driven Policies”. IEEE Transactions on Dependable and Secure Computing, 2024. DOI:10.1109/TDSC.2024.3356811 

2. Ahmed Akhtar, Basit Shafiq, Jaideep Vaidya, Ayesha Afzal, Shafay Shamail, and Omer Rana. “Blockchain Based Auditable Access Control for Distributed Business Processes.” 2020 IEEE 40th International Conference on Distributed Computing Systems (ICDCS), 2020, pp. 12-22., DOI:10.1109/ICDCS47774.2020.00015.