Javad Ghofrani

Ujjwal Patil, Javad Ghofrani

Deep reinforcement learning (DRL) has delivered strong results in domains such as Atari and Go, but it still suffers from high sample cost and weak transfer beyond the training setting. A common response is to reuse information from previously trained models through transfer, distillation, ensemble methods, or federated training instead of learning each target task from random initialization. The literature on these mechanisms is fragmented, and published comparisons are hard to interpret because tasks, baselines, and compute budgets differ. This paper presents a PRISMA-guided systematic review of empirical studies on pretrained knowledge reuse in DRL. Starting from 589 records retrieved from IEEE Xplore, the ACM Digital Library, and citation tracing, we screened 570 unique records and assessed 89 full texts. After applying the final eligibility criteria, 15 empirical studies remained in the main synthesis. We analyzed them qualitatively across three factors: source-target similarity, diversity among reused models, and the fairness of comparisons against from-scratch baselines. Three patterns recur across the surviving corpus. First, positive results are concentrated in settings where source and target tasks share substantial structure or where the method includes an explicit gating or alignment mechanism. Second, evidence for ensembles and federated aggregation is promising but sparse and mostly limited to narrow settings. Third, compute-matched comparisons are rare, which weakens claims about efficiency gains over stronger single-agent baselines. The paper contributes a narrower and internally consistent review scope, a study-level synthesis of empirical evidence, and a provisional independence spectrum that should be treated as a hypothesis for future benchmarking rather than a validated metric.

Javad Ghofrani, Ehsan Kozegar, Mohammad Divband Soorati et al.

Artificial Neural Networks (ANNs) replaced conventional software systems in various domains such as machine translation, natural language processing, and image processing. So, why do we need an repository for artificial neural networks? Those systems are developed with labeled data and we have strong dependencies between the data that is used for training and testing our network. Another challenge is the data quality as well as reuse-ability. There we are trying to apply concepts from classic software engineering that is not limited to the model, while data and code haven't been dealt with mostly in other projects. The first question that comes to mind might be, why don't we use GitHub, a well known widely spread tool for reuse, for our issue. And the reason why is that GitHub, although very good in its class is not developed for machine learning appliances and focuses more on software reuse. In addition to that GitHub does not allow to execute the code directly on the platform which would be very convenient for collaborative work on one project.

Javad Ghofrani, Paul Patoola, Daniel Richter et al.

Arrowhead Framework (AHF) is being developed to enable large-scale IoT based automation by providing an interoperability layer for local clouds. This framework aims to create an abstract model for distributed, heterogeneous, and non-linear systems. Managing the variability in such environments plays a key role in handling complex automation tasks such as in smart production systems. However, there is no standard solution available for handling the variability and configuration specifications in such environments. In this paper, we analyze the existing solutions for configuration management in industrial automation frameworks and provide leverage points for a standardization framework for handling configurations of automated production systems based on the concept of industrial internet of things.

Kirill Loisha, Javad Ghofrani, Dirk Reichelt

In the next few years, Blockchain will play a central role in IoT as a technology. It enables the traceability of processes between multiple parties independent of a central instance. Blockchain allows to make the processes more transparent, cheaper, and safer. This research paper was conducted as systematic literature search. Our aim is to understand current state of implementation in context of Blockchain Technology for digital protection of communication in industrial cyber-physical systems. We have extracted 28 primary papers from scientific databases and classified into different categories using visualizations. The results show that the focus in around 14\% papers is on solution proposal and implementation of use cases "Secure transfer of order data" using Ethereum Blockchain, 7\% papers applying Hyperledger Fabric and Multichain. The majority of research (around 43\%) is focusing on solution development for supply chain and process traceability.

Bastian Deutschmann, Javad Ghofrani, Dirk Reichelt

Production plants today are becoming more and more complicated through more automation and networking. It is becoming more difficult for humans to participate, due to higher speed and decreasing reaction time of these plants. Tendencies to improve production systems with the help of cognitive systems can be identified. The goal is to save resources and time. This mapping study gives an insight into the domain, categorizes different approaches and estimates their progress. Furthermore, it shows achieved optimizations and persisting problems and barriers. These representations should make it easier in the future to address concrete problems in this research field. Human-Machine Interaction and Knowledge Gaining/Sharing represent the largest categories of the domain. Most often, a gain in efficiency and maximized effectiveness can be achieved as optimization. The most common problem is the missing or only difficult generalization of the presented concepts.

Javad Ghofrani, Robert Kirschne, Daniel Rossburg et al.

Machine vision is critical to robotics due to a wide range of applications which rely on input from visual sensors such as autonomous mobile robots and smart production systems. To create the smart homes and systems of tomorrow, an overview about current challenges in the research field would be of use to identify further possible directions, created in a systematic and reproducible manner. In this work a systematic literature review was conducted covering research from the last 10 years. We screened 172 papers from four databases and selected 52 relevant papers. While robustness and computation time were improved greatly, occlusion and lighting variance are still the biggest problems faced. From the number of recent publications, we conclude that the observed field is of relevance and interest to the research community. Further challenges arise in many areas of the field.