ParticleTensorAI® Components
ParticleTensorAI® combines an optical probe with a control unit configured for AI-based analysis. The probe detects time-dependent light-scattering signals from droplets or particles. The control unit acquires and buffers the detector data for processing as Particle Tensors.
The three current ParticleTensorAI® configurations use the LSS1, LSS2, or LSS4 optical probe together with the corresponding ZEON, IMEA, or CLEON control unit.
Optical Probes
The optical probe defines the measurement region and detects scattered light as droplets or particles pass through the illuminated area. Depending on the probe, one, two, or four detector channels provide signal data for the ParticleTensorAI® analysis.
- LSS1: optical probe used in the single-channel PTA1 configuration.
- LSS2: optical probe used in the two-channel PTA2 configuration.
- LSS4: optical probe used in the four-channel PTA4 configuration.
Control Units
The control unit acquires the detector signals and provides the data for ParticleTensorAI® processing. In PTA operation, the recorded signal sequences are buffered and evaluated using an AI-based approach.
- ZEON: control unit used with the LSS1 probe in PTA1.
- IMEA: control unit used with the LSS2 probe in PTA2.
- CLEON: control unit used with the LSS4 probe in PTA4.
ParticleTensorAI® Configurations
| Version | Optical probe | Control unit | Detector channels |
|---|---|---|---|
| PTA1 | LSS1 | ZEON | 1 |
| PTA2 | LSS2 | IMEA | 2 |
| PTA4 | LSS4 | CLEON | 4 |
Shared Hardware Platforms
PTA1 and PTA2 use the same basic hardware platforms as SprayQuantAI® SQA1 and SQA2. For ParticleTensorAI® operation, the calculation electronics are configured for the AI-based signal-sequence evaluation.
PTA4 uses the LSS4 and CLEON hardware platform also used for TSTOF measurements. The ParticleTensorAI® configuration applies its tensor-based AI evaluation to the recorded light-scattering signals.
How the Components Work Together
The optical probe records scattered-light signals. The control unit acquires and buffers those signals. ParticleTensorAI® then arranges the measurement data as Particle Tensors for analysis by an AI model developed for the defined measurement task.
The configuration determines the available detector channels. The properties that can be classified or estimated depend on the measurement data, reference information, and validation of the AI model for the application.