Diploma in IL/DAgger and Mixing with Real-Time MPC
Online postgraduate diploma of 800 hours in IL/DAgger and blending with real-time MPC: 8 modules covering imitation learning, DAgger and interactive variants, model predictive control, real-time MPC, imitation of MPC controllers, hybrid learning-MPC architectures, and real-time validation and deployment, with an assigned technical tutor.
Diploma in IL/DAgger and Mixing with Real-Time MPC — studying from Australia
Data, artificial intelligence and simulation are central to Australian industries including mining, banking, health, agriculture and government. Organisations use machine learning, digital twins and advanced analytics to improve productivity, asset performance and decision making. National science agency CSIRO, along with universities and a growing start-up ecosystem, drives research and applied innovation. Australia has also developed national AI ethics principles, making responsible design an important professional consideration. Engineers in this area typically work for technology firms, consultancies, research organisations and large enterprises.
Fees for Australia: 2,560 A$ (AUD), with instalments and employer sponsorship available.
Relevant bodies in Australia
Typical roles
Questions from Australia
Are SEIUM qualifications recognised in Australia?
SEIUM qualifications are private qualifications, not official state degrees. Recognition for registration or licensing is decided by bodies such as Engineers Australia and state regulators, so please check with them directly for your specific goal.
What language are the programmes taught in?
All programmes are taught in English.
Take the controls.
Run a real engineering trade-off from this field. Move the sliders, watch the model react in real time and beat the challenge.
What you will be able to do
Explain the fundamentals of imitation learning and distribution shift.
Implement DAgger and its variants with human and algorithmic experts.
Formulate constrained MPC controllers with stability guarantees.
Run non-linear MPC in real time with embedded solvers.
Approximate MPC laws with verifiable neural networks.
Design hybrid architectures with MPC-based safety layers.
Validate hybrid controllers in simulation and hardware-in-the-loop.
Who it's for
- Predictive control engineers looking to combine MPC with learned policies.
- Robotics engineers working on manipulation and mobile platforms.
- Embedded systems engineers who deploy controllers under tight latency budgets.
- Imitation learning researchers moving towards real-world control.
- Autonomous driving engineers responsible for planning and control.
- Controller validation engineers.
- Engineering or science graduates aiming to enter industrial robotics.
What you'll work with
- IEC 61508:2010 (functional safety of E/E/PE safety-related systems)
- ISO 26262:2018 (road vehicles functional safety)
- ISO 13849-1:2023 (safety-related parts of machinery control systems)
- ISO 10218-1:2025 (industrial robot safety)
- ISO/TS 15066:2016 (collaborative robots)
- ISO 21448:2022 (safety of the intended functionality, SOTIF)
- IEC 61131-3:2013 (PLC programming languages)
- CasADi
- acados
- OSQP
- ROS 2
- MATLAB Model Predictive Control Toolbox
8 modules
1.1Behaviour cloning as supervised learning1.2Distribution shift and compounding errors1.3Human and expert demonstration data1.4Teleoperation and trajectory capture1.5State and action representation1.6Multimodal policies and mixture models1.7Diffusion policies and action transformers1.8Success and imitation metrics1.9Limits of the suboptimal expert1.10Workshop: behaviour cloning on a simulated arm
2.1The DAgger algorithm by Ross, Gordon and Bagnell2.2No-regret guarantees2.3HG-DAgger with human intervention2.4SafeDAgger and selective queries2.5Algorithmic experts and simulation oracles2.6Dataset aggregation and balancing2.7Labelling cost and query budget2.8Inverse reinforcement learning and GAIL2.9Comparison of BC, DAgger and GAIL2.10Lab: DAgger with MPC as the expert
3.1Formulation of the finite-horizon problem3.2Linear and non-linear prediction models3.3State and actuation constraints3.4Cost function and weighting3.5Stability with terminal costs and sets3.6Robust MPC and trajectory tubes3.7Economic and tracking MPC3.8State estimation with the Kalman filter3.9Tuning the prediction horizon3.10Practical exercise: trajectory-tracking MPC with CasADi
4.1QP solvers with OSQP and qpOASES4.2Sequential quadratic programming and RTI4.3Warm-starting and solution reuse4.4Explicit MPC and critical regions4.5Code generation with acados4.6Computational budget and latency4.7Embedded platforms and numerical precision4.8Solver failures and fallback strategies4.9Measuring worst-case computation time4.10Lab: embedded non-linear MPC with acados
5.1Approximating the MPC law with neural networks5.2State-space sampling5.3Policies with projection onto constraints5.4Verification of the approximating network5.5Learning the terminal cost5.6Distilling policies to lightweight hardware5.7Approximation error and margins5.8Guarantees through safety filters5.9Latency comparison with the original MPC5.10Practical exercise: neural network imitating a vehicle MPC
6.1Learned policy as solver warm-start6.2MPC as a safety layer for the policy6.3Learning residual models for MPC6.4Learning-based tuning of MPC weights6.5MPC with Gaussian-process uncertainty6.6Supervised switching between controllers6.7Control barrier functions and CBF-QP6.8Arbitration based on network confidence6.9Multi-rate execution6.10Case study: mobile robot navigation with an MPC layer
7.1Closed-loop simulation and disturbance injection7.2Hardware-in-the-loop testing of the controller7.3Real-time operating systems and ROS 27.4Runtime constraint monitoring7.5Data logging for new DAgger iterations7.6Applicable functional safety requirements7.7Safety of cobots and industrial robots7.8Failure mode analysis of the controller7.9Validation documentation7.10Case study: deployment on a collaborative manipulator
8.1System definition and specifications8.2Design of the reference MPC8.3Generation of demonstrations8.4Training with iterative DAgger8.5Hybrid integration with a safety layer8.6Real-time latency testing8.7Validation in simulation and HIL8.8Analysis of results and limitations8.9Report and presentation8.10Final project: final review with the tutor
How you'll learn
100% online training on the SEIUM campus, with downloadable material, recorded classes and an assigned technical tutor. There are no fixed timetables: you progress at your own pace within the intake, so it fits around your job.
Each module combines theory, guided reading of the original standard and applied exercises. The working software is CasADi, acados, OSQP, ROS 2 and MATLAB Model Predictive Control Toolbox, in free, academic or evaluation versions. Testing is studied on real procedures: vehicle trajectory tracking in simulation, control of collaborative manipulators and hardware-in-the-loop testing of embedded controllers. Assessment is continuous by module, plus the final capstone project.
Standards covered: IEC 61508:2010, ISO 26262:2018, ISO 13849-1:2023, ISO 10218-1:2025, ISO/TS 15066:2016, ISO 21448:2022, IEC 61131-3:2013.
Build it end to end
Your final project follows the real engineering workflow, step by step, reviewed by your tutor.
Where this takes you
Requirements
- Prior technical training or professional experience in the sector. No university degree is required.
- An internet connection and a computer capable of running the programme software.
- English sufficient to read technical standards.
Ready to start?
Send your application today. Our admissions team replies within 72 hours with your study plan and funding options.
Frequently asked questions
The programme totals 800 hours over 6–8 months. You set your own pace within each intake, and your tutor helps you plan the workload around your job.
You work with CasADi, acados, OSQP, ROS 2 and MATLAB Model Predictive Control Toolbox, using free, academic or evaluation versions. Your tutor indicates the free alternative wherever one exists.
A SEIUM Institute diploma, awarded on passing the continuous assessment and the final project. It is SEIUM's own qualification, not an official state degree; its recognition depends on the programme completed and the regulations of each country.
Yes. Modules passed are recognised in a higher programme in the same area (postgraduate diploma or master's). Ask about the recognition policy when you enrol.
Prior technical training or professional experience in the sector (no university degree required), a computer able to run the programme software and English sufficient to read technical standards.
Start dates are set by SEIUM's faculty for each intake. Our admissions team will confirm the next available date when you apply.
Related programmes
Diploma in Multi-Stakeholder Governance and Risk Assessment
Online postgraduate diploma of 800 hours in multi-party governance and risk assessment, with 8 modules covering data and AI governance, multi-party ecosystems, the European legal framework,…
Diploma in V&V of Critical Models and Safety Cases
Online postgraduate diploma of 800 hours in V&V of critical models and safety cases: 8 modules covering verification and validation concepts, code and solution verification, validation…
Diploma in Value Metrics and Data Monetization
Online postgraduate diploma of 800 hours in data value metrics and monetisation, with 8 modules covering data economics, data value metrics, causal impact measurement, monetisation models,…
More in Data, AI, Simulation & Analytics
All Data, AI, Simulation & Analytics programmes → · How to apply · Fees & financing · Scholarships