ai

Artificial intelligence is fundamentally changing the way products are developed. We use it throughout the entire development chain—from requirements analysis to code generation and debugging to test case generation. This shortens development cycles and increases test coverage without relinquishing technical responsibility. In addition, we implement AI directly in the product: as edge AI on the microcontroller, in real time, and without a cloud connection.Translated with DeepL.com (free version)

get in touch

AS A SOLUTION AND SYSTEM PROVIDER, WE BRING YOUR IDEAS TO LIFE!
und bringen sie auf Wunsch direkt in Ihr Produkt.

01

AI in the Engineering Process

We deliver robust, reliable, and cost-effective hardware precisely tailored to your requirements.

learn more

02

AI in Software Development

Our developers offer our clients more than just code. Because good software simplifies complex devices and systems.

learn more

03

AI in Testing and Quality Assurance

Firmware gives your hardware intelligence and brings it to life. We develop individual firmware solutions and program custom applications.

learn more

04

Solution and System Development

Our systems integrate seamlessly into your existing infrastructure and work together perfectly. This allows us to achieve maximum performance, efficiency, and reliability.

learn more

AI that delivers results in your project

Through the perfect interplay of functional hardware, suitable firmware, and modular software, we transform complex challenges into simple solutions.
start inquiry
1

AI in the Engineering Process

A significant portion of development effort is spent on structured, repetitive tasks. This is precisely where AI delivers measurable value: by checking requirements for completeness and consistency, maintaining checklists and documentation, and automating recurring process steps.
The result is shorter lead times with consistent diligence and more capacity for actual engineering work.
- Requirements engineering: checking for completeness, consistency, and testability
- Deriving acceptance criteria from requirements
- Generating project-specific review and inspection checklists
- Creating and maintaining technical documentation
- Automating recurring process steps
2

AI in Software Development

AI delivers the most value where there is clear structure and frequent repetition: in generating recurring code structures such as driver skeletons and interfaces.
Technical assessment remains the responsibility of our engineers.
- Debugging: Analyzing logs and error patterns
- Exploring existing or undocumented codebases
- Refactoring, porting, and migration
- AI-assisted code reviews
3

AI in Testing and Quality Assurance

Deriving test cases from requirements is thorough but time-consuming work. As the scope of requirements grows, so does the risk of coverage gaps. To address this, we use tools developed internally at habemus! in our production environment. Every test case is professionally reviewed and approved.
- Test case generation from requirements
- Systematic identification of edge cases
- Prioritization of regression runs
- Analysis of test logs and error reports
4

Edge AI in products

Running a model on a microcontroller means working with limited memory and a tight energy budget. The benefits justify the effort: real-time response, offline operation, and no transmission of sensitive data to third parties.
We have implemented Edge AI in several client projects and support the entire chain from data acquisition and training to integration into your firmware.
- On-device inference on microcontrollers
- Voice control without cloud connectivity
- Gesture recognition based on RADAR or camera
- Predictive maintenance and anomaly detection
- Data acquisition, model training, and optimization
- TensorFlow Lite Micro, ONNX Runtime, STM32Cube.AI, etc.

case study

Measurement system: Voice and gesture control, no cloud required


‍Technologies: On-device inference, speech recognition, 3D gesture recognition, multi-touch HMI, voice output
‍
Challenge: When working with measurement technology, hands are rarely free: test setups, connections, and documentation take priority. A manufacturer wanted to make its devices touch-free, controllable via voice and gestures. Processing had to take place directly on the device, respond in real time, and remain reliable even in noisy environments.
‍
‍Solution: We implemented speech recognition as edge AI directly on the device. Multiple microphones capture commands, and custom-developed algorithms translate them immediately into device functions. An integrated speaker provides feedback on measurement values and device status, allowing the operator to keep their eyes on the setup.
Additionally, we implemented touch-free gesture recognition that interprets hand movements in the immediate vicinity of the control interface. Both processes run without a network connection.
‍
Result: A measurement system that can be operated entirely touch-free, with voice output for measurement values and device status. Local processing makes it independent of network infrastructure and prevents the transmission of measurement data to third parties.