FreeScopes AI I - AI-Based Radar Analysis & Training | SkyRadar

FreeScopes AI I

FreeScopes AI I introduces trainees to artificial intelligence for classification and anomaly detection using radar, simulator or recorded data. It connects conventional radar signal processing with the practical steps required to select and label data, configure visual neural-network workflows, train models and evaluate results.

Trainees work without writing code and learn how conventional radar processing and data-driven methods complement each other.

Designed for ATSEP training, defence-oriented radar education, universities and radar research laboratories.

Request information View technical description

See how a visual FreeScopes workflow connects labelled radar data with neural-network training, evaluation and model reuse.

Training Advantages

Connect Radar and AI — Understand how radar and signal measurements become input for AI-oriented analysis.
Prepare and Label Data — Select suitable files and assign labels for supervised exercises.
Build Models Visually — Connect neural-network layers without writing code.
Train and Evaluate — Examine accuracy, validation accuracy, loss and convergence.
Compare Architectures — Explore Dense, Conv1D and hybrid workflows.
Complement Conventional DSP — Understand AI as another method in the radar-processing toolbox.

Move from Signal Processing to Data-Driven Radar Analysis

Conventional radar-processing modules show how signals are transformed mathematically. AI I adds a data-driven workflow: trainees select and label suitable measurements, configure a neural-network architecture, train it and evaluate the resulting model.

AI does not replace FFT, CFAR, MTI, Doppler processing or tracking. It builds on suitable data produced by conventional radar and signal-processing workflows.

What Trainees Do

Select radar, simulator or recorded measurement files
Assign labels to supervised training classes
Build baseline Dense classification networks
Compare shallow, deep and Conv1D-based architectures
Use Dropout and Batch Normalization and observe training behaviour
Configure epochs, batch size and validation split
Evaluate accuracy, validation accuracy, loss, convergence and overfitting
Save trained models and reload them for later evaluation or reuse

Main Functions

Data Preparation

AI Train Data Setup for selecting files and assigning labels
AI Input, Reshape and Flatten layers

Visual Model Building

Dense and Conv1D layers
MaxPool1D and GlobalAveragePooling1D
Batch Normalization and Dropout

Training and Evaluation

AI Model Compile and AI Model Train
AI Model Evaluate
Accuracy, validation accuracy, loss and convergence review

Model Storage and Reuse

AI Model Save
AI Trained Model for loading previously saved models
AI Model End

Where It Fits

Radar & DSP Foundation → Radar Data → FreeScopes AI I

AI I is an AI-oriented extension from the radar-processing foundation. It does not replace or extend the linear Basic I → Basic II → ATC I → ATC II surveillance-processing path.

Prerequisites and Compute Boundary

FreeScopes AI environment
A compatible radar, simulator or recorded/replayed dataset
Labelled training data
Suitable compute resources

Advanced exercises depend on data volume and server performance. The exact server requirement for larger AI exercises remains delivery-specific; AI I should not be understood as including separate server hardware.

Who It Is For

ATSEP instructors and technical training centres
Military radar academies
Universities and engineering programmes
Radar, signal-processing and AI research laboratories

Learning Outcomes

Trainees can explain how labelled radar data supports supervised learning; assemble and compare visual neural-network workflows; interpret accuracy, validation and loss behaviour; identify overfitting; and describe how AI-based analysis complements conventional radar DSP.

Related Products

Connect Radar Data with Practical AI Training

Use FreeScopes AI I to move from labelled measurements to visual model building, training and evaluation.

Request information