Dr. @Amrit Pal is an Associate Professor at Vellore Institute of Technology Chennai (India) working in Machine Learning, Deep Learning, Federated Learning, Edge Computing, and IoT. His research focuses on building efficient AI systems for real-world applications. In addition, he guides students in developing distributed and edge-adaptive models, and has worked on consultancy and sponsored projects, contributing to practical technology solutions. Besides his academic and research work, he enjoys playing badminton in his free time. 

He joined the Editorial Board of Scientific Reports in early 2024 and currently serves as Guest Editor for the Collection “Deep learning for real-time object detection”.

***

We recently invited Dr Pal for a Q&A where they answered questions about their research and what it is to work as an Editorial Board Member at Scientific Reports. Some of the questions they answered are below:

What do you like most about being an Editorial Board Member for Scientific Reports?  

What I value most about serving as an Editorial Board Member for Scientific Reports is the chance to contribute meaningfully to the global research community. I appreciate the opportunity to review high-quality manuscripts across diverse fields and help ensure that robust, impactful science is communicated effectively. The role enables me to support authors, maintain strong publication standards, and stay engaged with emerging research trends. Above all, it is rewarding to play a part in promoting transparent, influential, and openly accessible scientific knowledge. 

You are leading one of our Guest-Edited Collections, “Deep learning for real-time object detection”. What is your Collection focused on? And how it fits within the UN's sustainable development goals? What are the breakthroughs in this field?  

The Collection “Deep learning for real-time object detection” focuses on fast, accurate, and efficient visual recognition models that work reliably in real-world settings. It highlights advances in lightweight architectures, edge-AI deployment, and optimized detection frameworks for applications such as autonomous systems, smart cities, healthcare imaging, and environmental monitoring. The theme aligns with key UN Sustainable Development Goals, including SDG 3, 9, 11, and 13, by supporting safer mobility, improved diagnostics, sustainable infrastructure, and climate-aware monitoring. Recent breakthroughs include transformer-based detectors, efficient edge models, and event-driven vision systems that deliver high performance even on low-power devices. 

How do you think Scientific Reports supports UN's sustainable development goals? Do you think we could do more here and how?   

Scientific Reports supports the UN’s Sustainable Development Goals by making high-impact findings freely available. The journal enables wider adoption of innovations that can improve healthcare, strengthen infrastructure, protect ecosystems, and advance responsible technological growth. 

But yes, we can certainly do more. Encouraging more SDG-focused Collections, inviting experts working on sustainability topics, and highlighting interdisciplinary studies can further strengthen our impact. Providing clearer visibility to SDG-related research and promoting collaborations across regions would also help accelerate progress toward these global goals. 

Leaning on your expertise in image analysis algorithms, please give a piece of advice on how to approach/assess this type of study, highlighting strengths and limitations of the approach, particularly concerning the specific research question and methodology. 

Drawing on my experience in image analysis algorithms, I begin by assessing whether the research question is clearly defined and appropriately matched to the imaging modality and chosen computational approach. I look for strong data quality, reliable annotations, and a transparent methodological pipeline, including preprocessing, model design, training, and validation. Strengths often include innovative architectures, well-justified features, and solid comparisons with baseline methods. However, I remain cautious about small or biased datasets, overfitting, data leakage, and limited generalizability. I also consider interpretability, robustness, and whether the study discusses practical constraints and failure cases to ensure a balanced and credible evaluation. 

What are the biggest challenges that you see for the future of research and research dissemination? 

I think the major challenges ahead include ensuring authenticity as AI-generated content grows and maintaining strong standards of reproducibility. I also believe achieving sustainable open access and equitable visibility for all researchers will be essential. 

What are responsible ways to integrate AI into the peer review process, and how can editors feel empowered to explore these new tools to provide helpful solutions for authors, editors, reviewers, and AI experts? 

I believe the responsible integration of AI into the peer review process requires transparency, strong human oversight, and clear ethical guidelines to ensure it remains a supportive tool rather than a replacement for human expertise. I also feel that editors can be empowered through collaboration, training, and an understanding of how AI can automate routine tasks to enhance efficiency and fairness for everyone involved. 

Scopus: https://www.scopus.com/authid/detail.uri?authorId=57193491090 

ORCID: https://orcid.org/0000-0002-0555-9087