Specify & analyze
Task models, deadlines, scheduling policies, response times, and resource sharing.
2026–2027 · Computer Engineering
Real-Time Systems
Designing systems that deliver the right result at the right time, from real-time scheduling to verified and predictable AI-enabled systems.
Course idea
Correctness includes time
A real-time system must produce logically correct results within its timing constraints. This course develops task models, scheduling and schedulability analysis, resource-sharing protocols, and real-time operating-system concepts.
Timed automata and UPPAAL connect requirements to formal verification and model-based testing. Timing experiments and an integrated project extend these foundations to Edge AI, runtime assurance, and LLM-enabled cyber-physical systems. Students learn to distinguish model guarantees from measured evidence and explain the assumptions behind each.
Learning journey
From timing requirements to system evidence
Implement & observe
RTOS mechanisms, nano-kernel code, timing experiments, and platform interference.
Model & verify
Timed automata, UPPAAL properties, counterexamples, and model-derived tests.
Integrate & justify
Edge AI, timing budgets, runtime monitoring, validation, fallback, and project evidence.
Weekly structure
Course schedule
Week 9 is reserved for the midterm; Week 14 for project presentations and discussion.
| Week | Main topic | Core concepts |
|---|---|---|
| 1 | Introduction to Real-Time Systems | Logical and temporal correctness; hard, firm, and soft deadlines; latency, jitter, and WCET |
| 2 | Real-Time Task Models and Scheduling | Task models; static schedules and cyclic executives; cooperative and preemptive execution; RM, DM, EDF |
| 3 | Schedulability and Response-Time Analysis | Utilization tests; Liu–Layland bound; response-time analysis; blocking; harmonic task sets |
| 4 | Resource Sharing and Real-Time Concurrency | Critical sections; mutexes and semaphores; priority inversion; inheritance and ceiling protocols; deadlock |
| 5 | Real-Time Operating Systems | Tasks, interrupts, timers, IPC, and context switches; FreeRTOS/Zephyr; RTAI; nano-kernel examples |
| 6 | Timing in Real Computer Systems | Measured execution time versus WCET; caches, memory, OS interference, and multicore contention |
| 7 | Timed Automata for Real-Time Systems | Formal definitions and semantics; clocks, guards, invariants, and channels; networks; reachability; state explosion |
| 8 | Model Checking of Real-Time Systems | UPPAAL; reachability, safety, liveness, and leads-to properties; counterexamples; model-based testing |
| 9 | Midterm Examination — No Lecture | Examination week; no new lecture content |
| 10 | Edge AI Systems and Inference under Real-Time Constraints | End-to-end pipelines; CPU/GPU/NPU; timing and resource budgets; tail latency; model optimization |
| 11 | Scheduling and Resource Management for Edge AI | Mixed AI/RT workloads; shared accelerators; batching; overload; deadline-aware execution |
| 12 | Safe and Predictable AI-Enabled Real-Time Systems | Monitors, acceptance testers, watchdogs, and fallback; selected SPARC runtime-assurance and evidence ideas |
| 13 | LLMs in Real-Time and Cyber-Physical Systems | Variable latency; structured outputs; validation; bounded retries; safe fallback; end-to-end integration |
| 14 | Project Presentations and Discussion | Demonstrations, design decisions, verification, measurements, failure handling, and limitations; no new lecture unit |
Practice
Practical work & project
Guided exercises and one integrated project connect analysis, formal models, implementation, and experimental evidence.
Scheduling & timing
Build schedule traces, analyze response times and blocking, and investigate execution-time variability on real platforms.
Model → verify → test
Use timed automata and UPPAAL to check requirements, interpret counterexamples, and derive tests for implemented systems.
Design → integrate → defend
Develop timing and resource budgets in Weeks 10–11, integrate runtime assurance and experiments in Weeks 12–13, and present in Week 14. The project brief will specify deliverables and deadlines.
Assessment
Grading
Textbook
Primary reference
Real-Time Systems: Design Principles for Distributed Embedded Applications
Hermann Kopetz · Second edition · Springer · 2011
Course materials
Distribution
Lecture slides, source code, UPPAAL models, exercises, project instructions, announcements, and supplementary readings will be shared through the course learning environment. Assessment policies, learning outcomes, and the GenAI usage policy are available in the syllabus.
“How can we build systems that deliver the right result at the right time, and provide evidence for that claim?”
This is the central question revisited throughout the semester.