Machine Learning Engineer– Sensor Technology

Machine‑Learning Engineer – Eddy‑Current Sensor Systems (Industrial Sensors)
Location: [City, Country] – Hybrid / Remote options available
Employment type: Full‑time, permanent

---

About the Company

We are a fast‑growing, innovation‑driven technology firm that designs and manufactures advanced sensor solutions for the construction and infrastructure markets. Our flagship products combine radar, eddy‑current, and ultrasonic technologies to detect and characterize embedded steel reinforcement in concrete. With a strong portfolio of patents and a global customer base, we are now expanding the capabilities of our eddy‑current platform through state‑of‑the‑art machine‑learning (ML) techniques.

---

Role Overview

We are looking for a Machine‑Learning Engineer who will lead the development of a neural‑network‑based solution to dramatically improve metal‑detection accuracy and thickness estimation in reinforced concrete. You will work closely with hardware engineers, signal‑processing specialists, and software developers to turn raw sensor data into reliable, production‑ready predictions that can be deployed on our existing C‑based software stack.

---

Key Responsibilities

Area	What You’ll Do
Domain Understanding	• Acquire deep knowledge of our radar, eddy‑current, and ultrasound sensor hardware and signal‑processing pipelines. • Identify the physical and algorithmic limitations of the current eddy‑current system, especially for shallow reinforcement (first concrete layer).
Model Analysis & Design	• Review and benchmark existing analytical models for estimating embedded steel diameter. • Design, prototype, and iterate on PyTorch‑based neural networks (e.g., CNNs, RNNs, graph‑nets) that fuse multi‑modal sensor data.
Data Engineering	• Clean, augment, and label large‑scale sensor datasets (raw waveforms, derived features, ground‑truth measurements). • Implement robust data pipelines (PyTorch DataLoaders, Dask/Polars) for reproducible training/validation.
Training & Validation	• Conduct systematic hyper‑parameter sweeps, cross‑validation, and uncertainty quantification. • Develop custom loss functions that penalize thickness‑estimation errors in the critical near‑surface region.
Software Integration	• Export trained models to ONNX/TorchScript and embed them into the existing C library via a lightweight inference runtime (e.g., libtorch, TensorRT, or custom C‑API). • Write clean, well‑documented C/C++ wrappers and unit tests to ensure deterministic behavior on target hardware.
Performance & Deployment	• Optimize inference latency and memory footprint for real‑time operation on embedded processors (ARM Cortex‑M, DSP). • Create CI/CD pipelines (GitHub Actions, Jenkins) that automatically rebuild, test, and package the ML component with each firmware release.
Collaboration & Knowledge Transfer	• Partner with product managers to translate market requirements into technical specifications. • Mentor junior engineers and document best‑practice guidelines for ML‑enhanced sensor processing.

---

Required Qualifications

Skill	Minimum Requirement
Education	MSc or PhD in Electrical Engineering, Computer Science, Applied Physics, or a related field (or equivalent industry experience).
Machine Learning	3+ years of hands‑on experience building and deploying deep‑learning models with PyTorch (training, quantization, export to ONNX/TorchScript).
Signal Processing	Strong understanding of eddy‑current, radar, and ultrasonic signal fundamentals; ability to interpret raw waveform data.
Programming	Proficient in Python (data science stack) and C/C++ (integration, embedded development).
Model Deployment	Experience embedding neural networks into C‑based environments, preferably using libtorch, TensorRT, or similar runtimes.
Data Handling	Skilled at large‑scale data cleaning, feature engineering, and building reproducible pipelines (e.g., Pandas/Polars, Dask, PyTorch DataLoader).
Testing & CI	Familiar with unit testing (GoogleTest, pytest) and CI/CD for firmware/ML code.
Communication	Ability to explain complex technical concepts to cross‑functional teams and write clear documentation.

---

Preferred (Nice‑to‑Have) Skills

• Knowledge of finite‑element modeling or analytical eddy‑current formulations.
• Experience with edge‑AI toolchains (e.g., Arm CMSIS‑NN, TensorFlow Lite Micro).
• Prior work in construction‑tech, non‑destructive testing, or civil‑infrastructure monitoring.
• Publications or patents related to sensor‑fusion or ML for material characterization.

---

What We Offer

• Competitive salary + performance‑based bonus.
• Stock‑option plan – become a co‑owner of a high‑growth tech company.
• Flexible working hours and remote‑work allowance.
• State‑of‑the‑art lab facilities (hardware test rigs, high‑speed data acquisition).
• Continuous learning budget (conferences, courses, certifications).
• Collaborative, multidisciplinary team culture focused on impact and innovation.

---

How to Apply

1. Resume/CV – Highlight relevant ML projects, sensor‑fusion work, and any C/C++ integration experience.
2. Cover Letter – Briefly describe why you are excited about improving concrete‑reinforcement detection and how your background aligns with the responsibilities above.
3. Portfolio (optional) – Links to GitHub repos, papers, or demos that showcase your work with PyTorch, signal processing, or embedded inference.

Please send the above materials to hr@innovativesensors.com with the subject line “ML Engineer – Eddy‑Current Sensors – [Your Name]”.

We look forward to building the next generation of intelligent sensor solutions together!