PACE: A Personalized Adaptive Curriculum Engine for 9-1-1 Call-taker Training

Zirong Chen; Hongchao Zhang; Meiyi Ma

cs.AIMar 5, 2026

PACE: A Personalized Adaptive Curriculum Engine for 9-1-1 Call-taker Training

Authors: Zirong Chen, Hongchao Zhang, Meiyi Ma

Organizations: College of Connected Computing, Vanderbilt University, Nashville, Tennessee 37235, USA

Abstract

9-1-1 call-taking training requires mastery of over a thousand interdependent skills, covering diverse incident types and protocol-specific nuances. A nationwide labor shortage is already straining training capacity, but effective instruction still demands that trainers tailor objectives to each trainee's evolving competencies. This personalization burden is one that current practice cannot scale. Partnering with Metro Nashville Department of Emergency Communications (MNDEC), we propose PACE (Personalized Adaptive Curriculum Engine), a co-pilot system that augments trainer decision-making by (1) maintaining probabilistic beliefs over trainee skill states, (2) modeling individual learning and forgetting dynamics, and (3) recommending training scenarios that balance acquisition of new competencies with retention of existing ones. PACE propagates evidence over a structured skill graph to accelerate diagnostic coverage and applies contextual bandits to select scenarios that target gaps the trainee is prepared to address. Empirical results show that PACE achieves 19.50% faster time-to-competence and 10.95% higher terminal mastery compared to state-of-the-art frameworks. Co-pilot studies with practicing training officers further demonstrate a 95.45% alignment rate between PACE's and experts' pedagogical judgments on real-world cases. Under estimation, PACE cuts turnaround time to merely 34 seconds from 11.58 minutes, up to 95.08% reduction.

Figures & tables

Figure 1: Overview of PACE’s involvement in 9-1-1 training.

Figure 2: PACE runtime under adaptive phase after cold start.

		Fast Learner ( $\lambda$ =0.12, $\psi$ =0.15)					Moderate Learner ( $\lambda$ =0.07, $\psi$ =0.25)
Group	Method	C@10 $\uparrow$	C@30 $\uparrow$	C@50 $\uparrow$	Z2H $\downarrow$	RE $\uparrow$	C@10 $\uparrow$	C@30 $\uparrow$	C@50 $\uparrow$	Z2H $\downarrow$	RE $\uparrow$
Heuristics	Round-Robin	42.24 $\pm$ 7.16	53.43 $\pm$ 5.57	66.11 $\pm$ 7.37	$\diagup$	63.14 $\pm$ 4.11	34.19 $\pm$ 8.21	47.21 $\pm$ 7.20	55.15 $\pm$ 8.16	$\diagup$	51.43 $\pm$ 6.44
Heuristics	Deficit-Driven	57.40 $\pm$ 4.14	55.22 $\pm$ 6.84	56.67 $\pm$ 8.17	$\diagup$	59.46 $\pm$ 6.41	49.50 $\pm$ 5.15	50.11 $\pm$ 7.13	48.00 $\pm$ 8.77	$\diagup$	45.47 $\pm$ 7.17
LLM-based	GraphRAG	59.56 $\pm$ 7.72	60.50 $\pm$ 8.18	61.44 $\pm$ 9.31	$\diagup$	60.22 $\pm$ 9.21	51.43 $\pm$ 7.82	48.91 $\pm$ 9.22	50.46 $\pm$ 8.91	$\diagup$	49.84 $\pm$ 9.64
	GenMentor	65.67 $\pm$ 9.55	79.41 $\pm$ 8.33	84.15 $\pm$ 8.08	29.40 $\pm$ 6.60	86.87 $\pm$ 6.71	66.14 $\pm$ 10.11	73.34 $\pm$ 8.78	75.44 $\pm$ 6.41	$\diagup$	78.12 $\pm$ 8.83
	Agent4Edu	68.14 $\pm$ 8.27	82.06 $\pm$ 6.58	88.18 $\pm$ 5.57	27.58 $\pm$ 5.24	90.47 $\pm$ 5.12	62.41 $\pm$ 8.10	72.59 $\pm$ 9.66	84.57 $\pm$ 6.18	48.10 $\pm$ 1.90	86.17 $\pm$ 5.21

Table 1: System-level comparison with PACE ablations. C@ t (%) denotes coverage at session t . Z2H (zero-to-hero) denotes sessions to first competent episode. Slash means never observed, excluded for average calculation. RE (random exams) denotes held-out exam score.

Figure 3: Coverage progression under different belief granularities.

Figure 4: Convergence of approximation gap $\delta$ (scatter points, left axis) and belief variance $\bar{\sigma}^{2}$ (gray line $\pm$ std, right axis) across sessions. Point color indicates explore:exploit ratio within each batch ( $K$ =5): green denotes exploration-heavy sessions, red denotes exploitation-heavy sessions.

Personalized Learning

See all 18 papers

Oct 26, 2026

AdaTutor: Adaptive Tutoring through Pedagogical Strategy Selection

Effective tutoring requires determining not only what information to provide, but also which form of assistance to deliver. A tutor may address the same student error using various pedagogical strategies, adapting its approach as the interaction progresses. We introduce Comparative Pedagogical Supervision (CPS), a training framework to learn to select the optimal strategy and feedback for student learning. We evaluate CPS on math and science reasoning datasets. After five tutoring turns, CPS outperforms the strongest fixed-strategy baselines. The learned tutor transfers across datasets, and its strategy selection can also inform a separate API-based tutor.

Rajeev Bhatt Ambati, Shashank Srivastava, Snigdha Chaturvedi

UNC Chapel Hill

Intelligent Tutoring Systems Adaptive Learning Educational Feedback Generation

Ask PDF

Jun 23, 2026

Spatio-Temporal Retrieval-based Priors for Adaptive Computational Teaching in Driving

Learning-based automated coaching systems for complex motor tasks such as high-performance driving remain limited in the ability to be adaptive by their reliance only on local, context-dependent reasoning, failing to account for the long-term temporal nature of student learning and the cumulative impact of repeated teacher-student interactions. In this paper, we propose an imitation learning based computational model for adaptive teaching with a dedicated temporal reasoning module that can reason over the interaction history under low-data regimes. To compensate for limited amounts of interactive training data, and based on the repetitive nature of the teaching process, the model relies on a nearest neighbor retrieval and cross attention prior, reasoning only on a narrowed-down set of semantically similar past interactions with an encoder-decoder based concurrent teaching model. We validate our approach with (i) a novel semi-synthetic closed-loop longitudinal student-teacher interaction dataset based on Waymo Open Motion Dataset and (ii) a small-scale real-world naturalistic simulator race coaching dataset. Our results reveal the consistent advantage of our adaptive teaching model with the nearest neighbor retrieval and cross-attention prior over a non-adaptive baseline as well as a suite of adaptive models that differ in their choice of priors and temporal fusion mechanisms.

Deepak Edakkattil Gopinath, Xiongyi Cui, Jonathan DeCastro +2

Toyota Research Institute, Cambridge, MA, USA

Autonomous Driving Imitation Learning Teacher-Student Learning Adaptive Learning

Ask PDF

Sep 19, 2026

Adaptive Scaffolding Needs Contingency: An AI Tutor That Escalates and Fades on What the Learner Does

Coding assistants raise task performance, but learners plan and monitor less. Giving less away, the usual fix, conflates two things: how much work a system carries (cognitive load) and what the learner must decide before help arrives (metacognitive demand). Our principle, preserved metacognitive demand, holds the second constant and lets the first vary. CoMeT implements it: support rises when a learner fails at a decision point and fades on take-up. Within subjects, 131 adult learners used CoMeT, an unrestricted assistant and a question-only tutor on three Python tasks. CoMeT matched the question-only tutor's demand, delivered artifacts twice as often as the assistant, and frustrated learners less than the question-only tutor, with delegation and load unchanged. Learners often did not answer. Fading held when their turn addressed the decision under support, and CoMeT surrendered the full answer in one session in sixteen, against one in six for the question-only tutor.

Xinmeng Hou, Yuxuan Weng, Chin Hsien Yeh +5

National Institute of Education, Nanyang Technological University Singapore, Singapore · School of Computer Science and Engineering, Nanyang Technological University Singapore, Singapore · School of Art, Design and Media, Nanyang Technological University Singapore, Singapore +2

Instructional Scaffolding Intelligent Tutoring Systems Human-AI Collaboration Programming Education Cognitive Offloading

Ask PDF

Abstract

Personalized Learning

Explore similar work

AdaTutor: Adaptive Tutoring through Pedagogical Strategy Selection

Spatio-Temporal Retrieval-based Priors for Adaptive Computational Teaching in Driving

Adaptive Scaffolding Needs Contingency: An AI Tutor That Escalates and Fades on What the Learner Does