Speech-first English tutors are increasingly expected to correct pronunciation, explain regional usage, and keep the interaction fast enough for rehearsal. These requirements make energy use, latency, and feedback fidelity part of the same design problem. We revise the edge-cloud tutor as a formally specified, edge-first system for five English varieties: American, British, Indian, Australian, and Canadian English. The system selects among semantic-cache reuse, local retrieval-augmented generation, local generation with pronunciation scoring, and cloud fallback by maximizing a quality-energy-latency utility under explicit constraints. Its semantic cache stores only validated, evidence-versioned feedback; its inference cache reuses hot weights and short-context key-value states; and its feedback composer is restricted to evidence selected from a curated variety knowledge base. A controlled prototype evaluation over matched five-minute tutoring sessions shows that the hybrid cached design reduces total session energy by 41.0% and median latency by 44.3% relative to a cloud-only baseline, while preserving near-cloud variety fidelity (94.2% versus 95.0%). Compared with the same hybrid pipeline without caching, it reduces energy by 16.3% and median latency by 16.6%. Human-rated feedback evaluation further shows higher evidence support and lower answer leakage than a cloud NLP feedback baseline. The results do not claim that edge-first tutoring is always best: edge-only remains the lowest-energy mode, but loses fidelity and diagnostic depth. The main contribution is a transparent operating point for real-time English-variety learning where energy, latency, privacy, and pedagogical quality are jointly reported rather than treated as separate afterthoughts.
- Open Access
- Article
Author Information
Received: 06 Mar 2026 | Revised: 15 May 2026 | Accepted: 30 May 2026 | Published: 24 Jun 2026
Abstract
Keywords
energy-efficient AI | edge-cloud collaboration | English varieties | semantic caching | retrieval-augmented generation | pronunciation assessment | mobile learning
References
- 1.
Lo, C.K.; Yu, P.L.H.; Xu, S.; et al. Exploring the Application of ChatGPT in ESL/EFL Education and Related Research Issues: A Systematic Review of Empirical Studies. Smart Learn. Environ. 2024, 11, 50. https://doi.org/10.1186/s40561-024-00342-5.
- 2.
Nguyen Hoang Mai Tram, N.; Nguyen, T.T.; Tran, C.D. ChatGPT as a Tool for Self-Learning English among EFL Learners: A Multi-Methods Study. System 2024, 127, 103528. https://doi.org/10.1016/j.system.2024.103528.
- 3.
Du, J.; Daniel, B.K. Transforming Language Education: A Systematic Review of AI-Powered Chatbots for English as a Foreign Language Speaking Practice. Comput. Educ. Artif. Intell. 2024, 6, 100230. https://doi.org/10.1016/j.caeai.2024.100230.
- 4.
Karatas, F.; Abedi, F.Y.; Gunyel, F.O.; et al. Incorporating AI in Foreign Language Education: An Investigation into ChatGPT’s Effect on Foreign Language Learners. Educ. Inf. Technol. 2024, 29, 19343–19366. https://doi.org/10.1007/s10639-024-12574-6.
- 5.
Dai, Y.; Wu, Z. Mobile-Assisted Pronunciation Learning with Feedback from Peers and/or Automatic Speech Recognition: A Mixed-Methods Study. Comput. Assist. Lang. Learn. 2023, 36, 861–884. https://doi.org/10.1080/09588221.2021.1952272.
- 6.
Amrate, M.; Tsai, P.-H. Computer-Assisted Pronunciation Training: A Systematic Review. ReCALL 2025, 37, 22–42. https://doi.org/10.1017/S0958344024000181.
- 7.
Mohsen, M.A.; Mahdi, H.S.; AlThebi, S.H.; et al. A Scientometric Study of Computer-Assisted Pronunciation Training in Second Language Acquisition: Technological Affordances and Research Trends. Humanit. Soc. Sci. Commun. 2025, 12, 438. https://doi.org/10.1057/s41599-025-04474-y.
- 8.
Sungkur, R.K.; Shibdeen, N. Pronunciation Trainer for Second Language Learning Using Generative AI. Int. J. Educ. Technol. High. Educ. 2025, 22, 64. https://doi.org/10.1186/s41239-025-00561-x.
- 9.
Yan, B.C.; Chen, B. An Effective Hierarchical Graph Attention Network Modeling Approach for Pronunciation Assessment. IEEE/ACM Trans. Audio Speech Lang. Process. 2024, 32, 3974–3985. https://doi.org/10.1109/TASLP.2024.3449111.
- 10.
Lin, C.-C.; Huang, A.Y.Q.; Lu, O.H.T. Artificial Intelligence in Intelligent Tutoring Systems toward Sustainable Education: A Systematic Review. Smart Learn. Environ. 2023, 10, 41. https://doi.org/10.1186/s40561-023-00260-y.
- 11.
Mao, Y.; Yu, X.; Huang, K.-B.; et al. Green Edge AI: A Contemporary Survey. Proc. IEEE 2024, 112, 880–911. https://doi.org/10.1109/JPROC.2024.3437365.
- 12.
Li, Z.; Yu, H.; Fan, G.; et al. Energy-Efficient Offloading for DNN-Based Applications in Edge-Cloud Computing: A Hybrid Chaotic Evolutionary Approach. J. Parallel Distrib. Comput. 2024, 187, 104850. https://doi.org/10.1016/j.jpdc.2024.104850.
- 13.
Zhu, X.; Li, J.; Liu, Y.; et al. A Survey on Model Compression for Large Language Models. Trans. Assoc. Comput. Linguist. 2024, 12, 1556–1577. https://doi.org/10.1162/tacl_a_00704.
- 14.
Ren, S.; Tomlinson, B.; Black, R.W.; et al. Reconciling the Contrasting Narratives on the Environmental Impact of Large Language Models. Sci. Rep. 2024, 14, 26310. https://doi.org/10.1038/s41598-024-76682-6.
- 15.
Montoya Benitez, A.O.; Suarez Sarmiento, A.; Macias Lopez, E.M.; et al. Optimization of Energy Consumption in Voice Assistants through AI-Enabled Cache Implementation: Development and Evaluation of a Metric. Technologies 2025, 13, 19. https://doi.org/10.3390/technologies13010019.
- 16.
Kim, H.; Lee, J.; Bahn, H. Rethinking I/O Caching for Large Language Model Inference on Resource-Constrained Mobile Platforms. Mathematics 2025, 13, 3689. https://doi.org/10.3390/math13223689.
- 17.
Li, Z.; Wang, Z.; Wang, W.; et al. Retrieval-Augmented Generation for Educational Application: A Systematic Survey. Comput. Educ. Artif. Intell. 2025, 8, 100417. https://doi.org/10.1016/j.caeai.2025.100417.
- 18.
Reddig, J.M.; Arora, A.; MacLellan, C.J. Generating In-Context, Personalized Feedback for Intelligent Tutors with Large Language Models. Int. J. Artif. Intell. Educ. 2025, 35, 3459–3500. https://doi.org/10.1007/s40593-025-00505-6.
- 19.
Yue, X.; Gao, X.; Qian, X.; et al. Adapting Pre-Trained Self-Supervised Learning Model for Speech Recognition with Light-Weight Adapters. Electronics 2024, 13, 190. https://doi.org/10.3390/electronics13010190.
- 20.
Dong, R.; Chen, J.; Long, Y.; et al. Enhanced Cross-Modal Parallel Training for Improving End-to-End Accented Speech Recognition. Speech Commun. 2025, 169, 103188. https://doi.org/10.1016/j.specom.2025.103188.
- 21.
Davies, M.; Fuchs, R. Expanding Horizons in the Study of World Englishes with the 1.9 Billion Word Global Web-Based English Corpus (GloWbE). Engl. World-Wide 2015, 36, 1–28. https://doi.org/10.1075/eww.36.1.01dav.
- 22.
Greenbaum, S. (Ed.) Comparing English Worldwide: The International Corpus of English; Clarendon Press: Oxford, UK, 1996.
- 23.
Ardila, R.; Branson, M.; Davis, K.; et al. Common Voice: A Massively-Multilingual Speech Corpus. In Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020), Marseille, France, 11–16 May 2020; pp. 4218–4222.
This work is licensed under a Creative Commons Attribution 4.0 International License.
Suite 4002 Level 4, 447 Collins Street, Melbourne, Victoria 3000, Australia
General Inquiries: info@sciltp.com