We present a large-scale survey of sustainability communication and motivation in robotics research. Our analysis covers nearly 50,000 open-access papers from arXiv's cs.RO category published between 2015 and early 2026. In this study, we quantify how often papers mention social, ecological, and sustainability impacts, and we analyse their alignment with the UN Sustainable Development Goals (SDGs). The results reveal a persistent gap between the field's potential and its stated intent. While a large fraction of robotics papers can be mapped to SDG-relevant domains, explicit sustainability motivation remains remarkably low. Specifically, mentions of sustainability-related impacts are typically below 2%, explicit SDG references stay below 0.1%, and the proportion of sustainability-motivated papers remains below 5%. These trends suggest that while the field of robotics is advancing rapidly, sustainability is not yet a standard part of research framing. We conclude by proposing concrete actions for researchers, conferences, and institutions to close these awareness and motivation gaps, supporting a shift toward more intentional and responsible innovation.
Figures & tables
Figure 1: The Sustainability Gap : Out of 50,000 analyzed papers, only 7% mention broader societal or ecological impacts, and only about 4% are explicitly motivated by sustainability-related goals.
Table 2
Figure 2: 2025 update of planetary boundaries. Figure adapted from publicly released material by the Stockholm Resilience Centre and collaborators ( Kitzmann et al., 2025 ) .
Figure 3: The 17 UN Sustainable Development Goals (SDGs), used as the reference framework for our motivation and alignment analysis.
Metric
Value
Model Used
DeepSeek-V3
DeepSeek’s Carbon Intensity Factor (CIF)
∼0.6 kgCO 2 e / kWh ( Jegham et al., 2025 , Table 1)
Data Volume
1.2 billion tokens exchanged
Prompt Number
∼50,000
Average Prompt Size
Input: ∼25k tokens, Output: ∼750 tokens
Estimated Energy Per Prompt
∼13 Wh ( Jegham et al., 2025 , Table 4)
Table 1: Estimated resource and emissions footprint of the large-scale analysis pipeline.
Figure 4: Total number of papers published in arXiv’s cs.RO category over time.
Figure 5: The ratio of papers mentioning impacts versus the one not mentioning any on the left. The right graph shows the zoomed-in breakdown of impact mentions. Note: categories can overlap, one paper can mention multiple .
Figure 6: Quarterly impact mentions in robotics papers. The top graph compares the total number of papers with those mentioning any of the impacts. The bottom graph shows the evolution of the ratio of papers mentioning the impacts and SDGs in time.
Figure 7: Number of SDG Aligned papers, explicitly SDG Motivated papers, and papers with no direct SDG relevance. The percentage value shows the ratio of papers classified as SDG Aligned per SDG (out of all the 50,000 papers). Note: one paper can align to multiple SDG topics.
Figure 8: Share of explicitly SDG Motivated papers relative to the total SDG Aligned papers per SDG topic.
Figure 9: Comparison of total papers versus those motivated by sustainability, with percentage trend line.
Appendix figures & tables10 assets
Supplementary material from the paper’s appendix.
Appendix
Feature
Example A: Purely Technical
Example B: Sustainability Driven
Title
Factorization of Rational Motions: A Survey with Examples and Applications
Adaptive Sensor Placement Inspired by Bee Foraging: Towards Efficient Environment Monitoring
The paper is a survey on the factorization of rational motions for mechanism synthesis, with examples including Bennett linkages and open chains. Its motivation is purely academic and technical within kinematics, with no mention of sustainability, social, or ecological impacts. The developed mathematical methods are relevant to advancing engineering and innovation (SDG 9.5), but this is not an explicit aim. There is no direct alignment with the specific robotic applications listed in the IFR proposals.
The paper is explicitly motivated by addressing climate change and environmental degradation through sustainable robotics for forest monitoring (SDG 13, 15). The developed hybrid ABC-Levy algorithm for optimizing sensor placement in swarms is a technological innovation relevant to SDG 9. The authors mention ’sustainable robotics’ and ’ecological’ impact but do not cite the UN SDGs. The technology aligns with IFR’s SDG 15 proposals for using robots to protect ecosystems via environmental monitoring, though it focuses on data gathering rather than active soil or chemical management.
Appendix
Table 2: Comparison of model behavior on a purely technical paper versus a sustainability-driven paper.
Figure 10: The figures shows the 2D semantic projection of the 48,163 papers based on their SDG Alignment. Each node on the graph represents a unique combination of SDGs, and its size corresponds to the number of papers sharing that specific combination. Colored overlays indicate the distribution of papers aligned to each specific goal (SDGs 2 through 9), alongside their absolute count and relative percentage. This visualization demonstrates a massive centralization of research: the overwhelming majority of papers are heavily concentrated in the center of the landscape, dominated by SDG 9 and SDG 11: almost 30,000 papers are concentrated in two nodes labeled as aligned with SDG 9 and SDG 9 & SDG 11 combinations. In contrast, research addressing other sustainability challenges is widely scattered across the periphery in much lower-volume clusters. Note: SDG 1 is excluded due to insufficient data.
Figure 11: Semantic landscape of robotics research across UN Sustainable Development Goals (Part 2 of 2). Continuing the global corpus projection from Part 1, colored overlays indicate the distribution of papers aligned to goals 10 through 17. As established, node size corresponds to the volume of papers sharing a specific SDG combination. This mapping visually reinforces the centralised structure of the field: the massive central hub is dominated by infrastructural goals, most notably SDG 11 (Sustainable Cities). The ecological and climate-focused research, with Climate Action (SDG 13), Life Below Water (SDG 14), and Life on Land (SDG 15) is divided into more isolated, and low-volume clusters scattered across around the center.
Figure 12: Small multiples bar chart comparing the distribution of papers across all 17 SDGs by model. Each chart displays the absolute number of papers classified as having No SDG relevance, being SDG Aligned or being explicitly SDG Motivated. The shared axes highlight the consistent distribution shape across all four models.
Figure 13: Percentage of papers classified as explicitly motivated by sustainability (SDG Motivated) across all SDGs, for 4 tested models.
Figure 14: The chart shows the percentage of papers identified as explicitly mentioning the UN SDGs, Sustainability, Ecological impact, and Social impact for each LLM. While minor variations in sensitivity exist (e.g., Llama-3.1-70B-Instruct and DeepSeek-V3 extracting slightly more Social impact mentions), the relative hierarchy and the near-total absence of UN SDG mentions remain consistent.
Model
CIP
Energy per prompt (Wh)
DeepSeek-V3
0.600
13.162
Llama-3.1-70B
0.287
19.183
Appendix
Table 3: Carbon Intensity of Providers (CIP) and energy consumption per prompt for models with available data ( Jegham et al., 2025 ) .
Figure 15: Ratio of classified impact mentions across 200 runs for both Example A(Purely technical) and Example B (Sustainability-Driven).
Figure 16: Ratio of classified SDG motivation and alignment across 200 runs for both Example A(Purely technical) and Example B (Sustainability-Driven). The above plot shows the distribution of SDGs that the papers were classified as motivated by (SDG Motivated). The bottom plot shows the distribution of SDGs that the papers were classified as relevant or aligned to (SDG Aligned).
Figure 17: Ratio of classified SDG motivation and alignment targets across 200 runs for both Example A(Purely technical) and Example B (Sustainability-Driven). The above plot shows the distribution of SDG targets that the papers were classified as motivated by (SDG Motivated). The bottom plot shows the distribution of SDG targets that the papers were classified as relevant or aligned to (SDG Aligned).
For most of human history, we have not thought systematically about how and why we incorporate aspects of the natural world into our designs. The lack of a systematic approach has resulted in inconsistencies in motivations and methods that make it difficult to predict or evaluate the success of bio-inspired design. This mismatch between expectations and results can lead to disappointment when a reader considers a bio-inspired design to be superficial, weak, or incomplete. This is especially true in the field of Robotics, in which similarity to a biological system might be the driving motivation for construction. In an effort to assist robotics researchers justify their specific bio-inspired approach and to assist funding program managers with discerning the value of different bio-inspired approaches, here we propose a taxonomy of motivations for bio-inspired design and describe the potential significant contributions that are likely to result from different approaches.
Margaret J. Zhang, Justin Ting, Talia Y. Moore
Mechanical Engineering, University of Michigan, USA 48109 · Electrical and Computer Engineering, University of Michigan, USA 48109 · Robotics, Mechanical Engineering, Ecology and Evolutionary Biology, Museum of Zoology, University of Michigan, USA 48109
Despite a surge in robotics research dedicated to inferring and understanding human intent, a universally accepted definition remains elusive since existing works often equate human intention with specific task-related goals. This article seeks to address this gap by examining the multifaceted nature of intention. Drawing on insights from psychology, it attempts to consolidate a definition of intention into a comprehensible framework for a broader audience. The article classifies different types of intention based on psychological and communication studies, offering guidance to researchers shifting from pure technical enhancements to a more human-centric perspective in robotics. It then demonstrates how various robotics studies can be aligned with these intention categories. Finally, through in-depth analyses of collaborative search and object transport use cases, the article underscores the significance of considering the diverse facets of human intention.
J. E. Domínguez-Vidal, Alberto Sanfeliu
Institut de Robòtica i Informàtica Industrial, Llorens i Artigas 4-6, Barcelona, 08028, Spain. · Universitat Politècnica de Catalunya, Jordi Girona, 31, Barcelona, 08034, Spain.
This paper takes issue with the recent themes of both the RO-MAN and the HRI conferences for their portrayal of a future human-robot society as inevitable. The focus is on discussing how such statements ultimately shape research. By treating a future human-robot society as a fait accompli, license is given for user studies to imagine any scenario they like, no matter whether it has any ecological relevance, and to emphasise the scenario design over actually creating robot abilities needed to fullfill the imagined role. Meanwhile, research that focusses on actual societal needs, without assuming that robots are a solution, is deprioritised, as is technical development, in particular with respect to abilities that are necessary to enable robots that function as social agents rather than a mere automation of tasks. A frame that simply assumes a robot future not only detracts from scientific advancement in favour of a techno-solutionism we ought to resist, it is also self-defeating as it risks stifling the research needed to bring it about. We should therefore reject attempts to frame and promote the field in terms of the inevitable social robot and instead focus on one that facilitates advances in the field regardless of what the future holds. This paper suggests that a renewed focus on cognitive mechanisms necessary for the "I" in HRI would be a good starting point.
Serge Thill
Donders Institute for Brain, Cognition, and Behaviour Radboud University Nijmegen 6525GD Nijmegen, The Netherlands