Jell AI

Jell AI Corp ยท Engineering consultancy

Real-time simulation and computer vision, from the engineers who built the tools.

A senior team with more than 75 years of combined experience in real-time simulation, computer vision and control systems. We were core developers of SIMulation Workbench, and we help automotive, aerospace and security teams design, integrate and debug systems where timing is not negotiable.

Talk to us What we do
75+ yearsCombined experience in real-time simulation, computer vision and control systems.
SimWB core developersWe were core developers of SIMulation Workbench, the real-time simulation framework from Concurrent Real-Time.
Automotive, aerospace, defense, public venuesDriving simulators, HIL rigs, missile and aircraft test benches, and threat assessment for crowded public spaces.

What we do

Two practices, one discipline: systems that have to respond on time.

Real-time simulation

Real-time simulation and hardware-in-the-loop

Architecture, I/O integration, model integration and performance tuning for driving simulators, HIL rigs and test benches in automotive and aerospace.

  • Real-time architecture: scheduling, multi-rate models, core allocation, latency and jitter budgets
  • I/O and bus integration: CAN, SOME/IP, analog, digital, PWM and the other interfaces listed below
  • Integration with Simulink Real-Time, dSPACE and NI systems, and with Simpack, MSC Adams, Simulink, rFpro and Diamond Visionics
  • Rig bring-up, troubleshooting and test automation
More on real-time simulation

Computer vision

Computer vision and edge AI for security

End-to-end architecture of camera, edge-compute and detection systems for threat assessment in soft targets such as airports, stadiums and train stations.

  • Camera and optics selection, placement and coverage
  • Edge computing hardware chosen and sized for the detection workload
  • Detection and tracking pipelines designed to a latency budget
  • Integration with the security operation that acts on the alerts
More on computer vision

Interfaces and tools

The buses, signals and tools a modern rig is built from.

Interface and tool support follows what SIMulation Workbench supports today. If your rig uses something that is not listed, ask.

Buses and protocols

  • CAN / CAN FD
  • FlexRay
  • Automotive Ethernet
  • SOME/IP
  • ARINC 429
  • MIL-STD-1553
  • RS-232 / 422 / 485
  • UDP / TCP
  • Reflective memory

Signals

  • Analog in / out
  • Digital I/O
  • PWM
  • Pulse / frequency
  • Resolver / synchro
  • LVDT / RVDT
  • Resistance simulation
  • Relays

Tools and platforms

  • SIMulation Workbench
  • MATLAB / Simulink
  • Simulink Real-Time
  • dSPACE
  • NI
  • Simpack
  • MSC Adams
  • FMI / FMU
  • rFpro
  • Diamond Visionics

Where we have applied it

Rigs we have built and debugged.

Driving simulators

Driver-in-the-loop systems where vehicle dynamics, motion cueing, image generation and cockpit I/O must stay in step, frame after frame.

Hardware-in-the-loop rigs

ECU and subsystem test benches with simulated sensors, actuators and bus traffic, fault injection and automated regression runs.

Missile testing

Guidance, navigation and control validation against real sensors and actuators, with deterministic timing and repeatable scenarios.

Aircraft systems test rigs

Including high-lift system testing for Airbus, where the rig has to reproduce the aircraft's interfaces, loads and timing faithfully.

Person detection boxes drawn over a crowded station platform
Person detection on a crowded platform

Computer vision

From camera to verdict, at the edge.

We architected the complete detection setup for threat assessment in soft targets: the cameras, the edge computing devices, the detection models and the way alerts reach the people who act on them.

  • Camera and lens selection for the coverage, resolution and lighting on site
  • Edge compute selected and sized for the models and the number of streams
  • Detection and tracking pipelines designed to a latency budget
  • Deployed in environments such as airports, stadiums and train stations
How we approach vision projects

How we work

From a short assessment to hands-on integration.

  1. Assessment

    A short, fixed-scope review of your rig, models and I/O that ends in a written report with specific recommendations.

  2. Architecture

    Timing budgets, core allocation, I/O mapping, model partitioning and tool selection, documented before anything is built.

  3. Integration

    We work alongside your engineers on the rig: wiring and signal mapping, model import, scripting and tuning, until it runs cleanly.

  4. Support

    Retained help for upgrades, new interfaces and the problems that only show up in the middle of a test campaign.

The team

Control engineers who spent their careers inside the simulation loop.

Jell AI is a small team with more than 75 years of combined experience. We come from control systems, and we spent much of that time building SIMulation Workbench and the rigs that ran on it: driving simulators, hardware-in-the-loop benches, missile test systems and aircraft systems test rigs. More recently we have applied the same discipline to computer vision for security at crowded public venues.

  • We built the scheduling, I/O and model-integration layers that rigs depend on, rather than only using them.
  • We work across the toolchain: SIMulation Workbench, Simulink Real-Time, dSPACE and NI on the real-time side; cameras, edge devices and detection models on the vision side.
  • We stay hands-on, from architecture through commissioning and handover.

Start a conversation

Tell us about your rig, your loop rate and your deadline.

Contact Jell AI