Postgraduate Diploma in Map-Based Prediction and Danger Hotspots
Online 800-hour Postgraduate Diploma in Map-Based Prediction and Danger Hotspots. 8 modules covering ISO 21448:2022 (SOTIF), ISO/PAS 8800:2024 and PyTorch. Assigned tutor and open enrolment.
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
Understand the role of prediction in autonomous driving.
Represent the map effectively for prediction.
Implement map-conditioned prediction models.
Identify danger hotspots using spatial methods.
Integrate risk priors into prediction models.
Evaluate the robustness of predictors.
Validate prediction modules against SOTIF criteria.
Who it's for
- Behaviour prediction engineer looking to specialise in Map-Based Prediction and Danger Hotspots.
- Mobility data scientist looking to specialise in Map-Based Prediction and Danger Hotspots.
- Machine learning engineer looking to specialise in Map-Based Prediction and Danger Hotspots.
- Road safety analyst looking to specialise in Map-Based Prediction and Danger Hotspots.
- Autonomous driving researcher looking to specialise in Map-Based Prediction and Danger Hotspots.
- Validation engineer looking to specialise in Map-Based Prediction and Danger Hotspots.
- Engineering or science graduates looking to enter the automotive sector.
What you'll work with
- ISO 21448:2022 (SOTIF)
- ISO/PAS 8800:2024
- ISO 34502:2022
- Directive (EU) 2019/1936
- Regulation (EU) 2016/679 (GDPR)
- UL 4600
- Regulation (EU) 2024/1689 (AI Act)
8 modules · 800 h
1.1The role of prediction in the autonomous stack1.2Prediction horizons and frequency1.3Vehicles, pedestrians and cyclists1.4Uncertainty and multimodality1.5Physical motion models1.6Manoeuvre-based models1.7Argoverse 2 and Waymo Motion datasets1.8minADE, minFDE and miss rate metrics1.9Marginal versus joint prediction1.10Case study: analysis of prediction errors
2.1Rasterisation of the map and context2.2VectorNet vector representation2.3LaneGCN lane graphs2.4Rules and traffic light states2.5Candidate routes and centrelines2.6Intersection context2.7Incomplete or erroneous maps2.8Computational cost2.9Prediction without a prior map2.10Lab: vector encoding of a map
3.1Recurrent networks with map context3.2Goal-based models: TNT and DenseTNT3.3Scene transformers such as Wayformer3.4Diffusion models for trajectories3.5Joint prediction of multiple agents3.6Attention between agents3.7Future occupancy prediction3.8Distillation for deployment3.9Autoregressive motion-token models3.10Lab: training on Argoverse 2
4.1Hotspots and concentration stretches4.2Empirical Bayes methods4.3Getis-Ord Gi* statistic4.4Spatiotemporal clustering4.5Surrogate events: hard braking and abrupt manoeuvres4.6Fleet data for detecting conflicts4.7Validating hotspots against crashes4.8Visualisation and communication4.9Aggregation by time band4.10Practical: hotspots using braking data
5.1Location-based risk priors5.2Adjusting uncertainty in hazardous areas5.3Anomalous behaviours5.4Models sensitive to local context5.5Planner attention to critical zones5.6Transfer between cities5.7Continuous updating of priors5.8Safety benefit5.9Risk-weighted training5.10Lab: prediction with hotspot priors
6.1Evaluation by scenario type6.2Long-tail cases6.3Robustness to map errors6.4Adversarial attacks on prediction6.5Probability calibration6.6Planning-oriented metrics6.7Closed-loop simulation6.8Comparison with baselines6.9Robustness to unseen agents6.10Case study: robustness report for a predictor
7.1SOTIF requirements for prediction7.2Critical scenarios derived from hotspots7.3In-service monitoring7.4Optimisation for embedded hardware7.5Explainability of predictions7.6Personal data in fleet analysis7.7Improvement with field data7.8Safety documentation7.9Latency and update rate7.10Workshop: validation plan for a prediction module
8.1Data and city selection8.2Danger hotspot analysis8.3Map representation8.4Baseline prediction model8.5Integrating risk priors8.6Stratified evaluation8.7Robustness analysis8.8Deployment proposal8.9Report and presentation8.10Final 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. No fixed timetables: students progress at their own pace within the intake, which makes it possible to combine study with work.
Each module combines theory, guided reading of the original standard and applied exercises. The working software is PyTorch, Argoverse 2 API, Waymax, QGIS, Python (PySAL), in free, academic or evaluation versions. Assessments are based on real training procedures for a map-conditioned predictor, hotspot detection from hard-braking events, and stratified evaluation of predictions at intersections. Assessment is continuous across each module plus the final integrative project.
Build it end to end
Your final project follows the real engineering workflow, step by step, reviewed by your tutor.
Understand the role of prediction in autonomous driving.
Represent the map effectively for prediction.
Implement map-conditioned prediction models.
Identify danger hotspots using spatial methods.
Integrate risk priors into prediction models.
Where this takes you
Requirements
- Prior technical training or professional experience in the sector. A university degree is not required.
- Internet connection and a computer capable of running the programme's software.
- Sufficient level of Spanish to follow technical regulatory texts.
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 spread over 6 to 8 months. That is an average of 10 to 15 hours per week, compatible with a full-time job.
You work with PyTorch, Argoverse 2 API, Waymax, QGIS, Python (PySAL). Free, academic or evaluation versions are used, and the tutor indicates the free alternative where one exists.
An own diploma from SEIUM on passing the continuous assessment and the integrative project. It is in-house training, not an official Ministry qualification.
Yes. Modules passed in a course or postgraduate diploma in the same area are recognised in the higher programme. Check the recognition policy when you enrol.
Related programmes
Postgraduate Diploma in V2X PKI, Latency and QoS for Road Safety
Online 800-hour Postgraduate Diploma in V2X PKI, Latency and QoS for Road Safety. 8 modules covering ETSI TS 103 097, ETSI TS 102 941 and Wireshark.…
Diploma in Local Regulations and Last Mile Operation
Online postgraduate diploma of 800 hours in local regulations and last-mile operation. 8 modules covering automated last-mile logistics, the legal framework for public roads, comparative law,…
Postgraduate Diploma in MPC and Real-Time Optimal Control for AD
Online 800-hour Postgraduate Diploma in MPC and Real-Time Optimal Control for AD. 8 modules covering ISO 26262:2018, ISO 11270:2014 and acados. Assigned tutor and open enrolment.
More in Autonomous, ADAS & Software-Defined Vehicle
All Autonomous, ADAS & Software-Defined Vehicle programmes → · How to apply · Fees & financing · Scholarships