Workshop · IEEE Big Data 2026

Big Data Analytics & AI for Intrusion Detection in Cyber-Physical Systems

BDACPS 2026

Catching the intrusion before the damage.

When an attacker reaches through a digital channel into a physical process, the signal shifts before anything breaks. This workshop is about detecting that shift — at the scale of real cyber-physical systems, with big data analytics, machine learning and signal processing.

Phoenix, Arizona · USA Dec 14–17, 2026 Full-day · Onsite
SENSOR_STREAM · OT/ICS ch:04
MONITORING
⚠ INTRUSION · flagged
baseline nominal Δ 4.2σ class false-data-injection status ANOMALY
01 · Why this workshop

Defending the systems where bits move atoms.

Cyber-physical systems now run the grid, the water, the factory floor and the operating room. The same connectivity that makes them efficient gives adversaries a path from a network packet to a physical consequence.

Conventional intrusion detection was built for IT — not for the timing constraints, proprietary protocols and safety-critical physics of operational technology. Meanwhile the data keeps growing: time-series sensor readings, industrial network traffic, operational logs and actuator states arriving in volume, velocity and variety that only big-data pipelines can absorb.

Deep learning has proven it can learn the behavioral fingerprints hidden in that data. BDACPS 2026 brings together the data-engineering, signal-processing, industrial-IoT and security communities to push detection from the lab into live production environments — and to confront the hard parts head-on.

Stuxnet. Triton. Florida, 2021.
Landmark attacks where the digital reached the physical — and detection arrived late.

Hosted under IEEE Big Data 2026's focus on scalable data infrastructure and intelligent analysis, this workshop extends that agenda into operational-technology security, building on the team's prior SPID-CPS @ IEEE ICASSP 2024 satellite event.

02 · Call for papers

Topics of interest.

We welcome original work — theoretical foundations and applied systems alike — across six tracks. The list is indicative, not exhaustive.

A — Architectures

Big data architectures for CPS security

  • Scalable ingestion & preprocessing for heterogeneous sensor streams
  • Real-time stream processing (Kafka, Flink, Spark Streaming)
  • Time-series & data-lake stores tuned for OT/ICS analytics
  • Edge–cloud continuum and fog approaches to distributed detection
B — Learning

Machine & deep learning for detection

  • Supervised, semi- and unsupervised methods for CPS anomalies
  • RNNs, CNNs and transformers for time-series security analysis
  • Graph neural networks for topology-aware anomaly detection
  • Generative models (GANs, VAEs, diffusion) for attack augmentation
  • Federated & privacy-preserving learning across industrial sites
C — Signals

Signal processing techniques

  • Frequency & time-frequency analysis (FFT, wavelets, EMD)
  • Change-point detection for attack-onset identification
  • Sensor fusion & multi-modal integration for robust monitoring
  • Physics-informed methods separating process drift from tampering
D — Trust

Explainability, robustness & trust

  • XAI for transparent decisions in safety-critical contexts
  • Adversarial robustness against evasion and poisoning
  • Uncertainty quantification & confidence calibration
  • Certification, verification and audit for regulated CPS
E — Benchmarks

Datasets, benchmarks & evaluation

  • New CPS security datasets (SWaT, BATADAL, HAI, iTrust…)
  • Evaluation metrics tailored to CPS detection performance
  • Digital twins & simulation for realistic attack scenarios
  • Anonymization & synthetic data preserving statistical fidelity
F — Threats

Emerging threats & attack vectors

  • Stealthy multi-stage attacks (APTs, false data injection, replay)
  • IIoT, smart grids, autonomous vehicles and smart cities
  • 5G-enabled CPS and time-sensitive networking (TSN)
  • Cross-domain threat intelligence & collaborative defense
03 · Important dates

Key deadlines.

All deadlines are 23:59 AoE. Workshop deadlines are tentative and will be aligned with the IEEE Big Data 2026 workshop schedule.

01 / SUBMIT
Oct 30, 2026
Paper submission deadline
tentative
02 / DECIDE
Nov 13, 2026
Notification of acceptance
tentative
03 / FINALIZE
Dec 1, 2026
Camera-ready & registration
tentative
04 / MEET
Dec 14–17
IEEE Big Data 2026 · Phoenix, AZ
confirmed
04 · Submission

Send us your work.

Papers are peer-reviewed and accepted contributions appear in the IEEE Big Data 2026 proceedings, indexed in IEEE Xplore.

  • FormatIEEE 2-column
  • Lengthup to 10 pages
  • Reviewdouble-checked, single track
  • SubmissionIEEE BigData system
  • ProceedingsIEEE Xplore
Open the submission system →
  • Oral presentations

    15–20 minutes including Q&A, in a single full-day track.

  • Invited keynotes

    Distinguished voices from academia and industry on CPS security at scale.

  • Open-challenges panel

    A moderated discussion on the unsolved problems in big-data-driven CPS defense.

  • Posters & demos

    An interactive showcase and networking session, subject to submission volume.

05 · Organization

Program chairs.

AG

Antonio Galli

Program Chair
University of Naples Federico II · Dept. of Electrical Engineering & IT

Assistant Professor working on deep learning and big data analytics, with a focus on AI for Industry 4.0. PhD in Technology, Innovation & Management; co-organizer of SPID-CPS @ IEEE ICASSP 2024.

AF

Antonino Ferraro

Program Chair · Member, IEEE
Pegaso University · Dept. of Information Science & Technology

Tenure-track Assistant Professor in time-series forecasting, data-driven AI and explainable AI applied to CPS and healthcare. TPC member for IJCAI 2026 and guest editor for several international journals.

VM

Vincenzo Moscato

Program Chair
University of Naples Federico II · Dept. of Electrical Engineering & IT

Full Professor of Database and Information Systems, author of 100+ journal and conference publications. Research spans multimedia, knowledge management and big data analytics.

Program committee
Michela Gravina
Univ. of Naples Federico II, IT
Gian Marco Orlando
Univ. of Naples Federico II, IT
Diego Russo
Univ. of Bergamo, IT
Flora Amato
Univ. of Naples Federico II, IT
Giancarlo Sperlì
Univ. of Naples Federico II, IT
Andrea Vignali
Univ. of Naples Federico II, IT
Martina Iammarino
Pegaso University, IT
Fabiano Pecorelli
Pegaso University, IT
Marco Postiglione
Northwestern University, USA
Valerio La Gatta
Northwestern University, USA
Angelo Lorusso
Pegaso University, IT
Lorenzo Malandri
Univ. of Milano-Bicocca, IT
Mouzhi Ge
Deggendorf Inst. of Technology, DE
Massimiliano Albanese
George Mason University, USA
Fabio Persia
Univ. of L'Aquila, IT