Computer Vision Engineer

Autoliv is seeking a Senior Industrial AI Engineer / Edge AI Technical Lead to serve as the primary technical authority for the Global AI Operations (GAIO) programme. This senior individual-contributor and technical leadership role combines deep, hands-on expertise in computer vision, machine learning, and MLOps with the architectural breadth required to guide end-to-end deployments across manufacturing sites worldwide. The position builds on our established Data Science and AI Engineer profile, with broader accountability for industrial edge systems and multi-site technical leadership. The role requires 10+ years of progressive experience. The successful candidate will operate as an Industrial Edge AI Technical Leader, combining deep AI and MLOps expertise with hands-on edge system and network diagnostics. The candidate must also have sufficient fluency in OT, Kubernetes, security, and infrastructure to lead deployments and engage specialists effectively. The role bridges advanced computer vision research and industrial-grade production systems, with ownership of the full lifecycle—from proof of concept to scalable, mission-critical, real-time quality inspection on the manufacturing floor.
Total Experience 10+ years of progressive, hands-on experience in software engineering and AI/ML systems, including at least 5 years in production-grade computer vision or industrial AI deployments. • At least 3 years of experience in a senior, principal, or technical lead role, with responsibility for architecture decisions, junior-engineer mentoring, and multi-team deployment coordination. • Demonstrated experience delivering AI systems in manufacturing, automotive, logistics, or similarly constrained industrial environments. Soft Skills & Leadership • Demonstrated ability to lead multi-team, multi-site technology deployments using formal readiness gates, clear RACI ownership, and structured risk management. • Strong communication skills, with the ability to write precise technical specifications for specialist teams and explain AI system concepts to plant operations, quality, and executive audiences. • Proven mentorship experience, including coaching junior engineers in ML practices, Linux diagnostics, structured troubleshooting, and incident documentation. • A strong bias toward independent diagnosis and evidence-based escalation, with the ability to operate effectively in complex, ambiguous plant environments without constant managerial direction. PREFERRED QUALIFICATIONS • Experience with PLC communication and integration across Allen-Bradley, Siemens, Omron, or Mitsubishi systems, including tag and state-machine concepts, trigger/heartbeat/result patterns, and protocol drivers. • OT integration experience, including OPC UA, MQTT broker configuration, SCADA/MES connectivity, and plant-historian integration. • Advanced Linux expertise, including network-proxy configuration, performance profiling with GCC/GDB, kernel-parameter tuning, and NUMA-aware memory management for GPU workloads. • Security expertise in TLS/mTLS certificate lifecycle management, managed identity and RBAC in Azure or an equivalent cloud platform, SIEM integration, and OT cybersecurity practices aligned with NIST SP 800-82. • Relevant certifications, such as NVIDIA Deep Learning Institute (Inference Optimization), CKA (Certified Kubernetes Administrator), RHCSA/RHCE, AWS or Azure AI/ML specialist certifications, or CompTIA Security+. • Experience with digital-twin platforms or simulation-based synthetic data generation for pre training vision models in new manufacturing contexts.
1. Computer Vision & Deep Learning — Technical Leadership • Lead the design, training, fine-tuning, and evaluation of state-of-the-art computer vision models for industrial quality inspection, including object detection, segmentation, anomaly detection, feature alignment, and OCR. • Own the full model lifecycle, from dataset strategy, labeling tools, and class-imbalance management to training pipeline design, hyperparameter optimization, and production performance governance. • Apply deep learning architectures (YOLO, Detectron2, SAM, and CLIP) and classical image processing techniques (OpenCV, spatial transformations, color-space manipulation, and morphological filtering) to solve manufacturing inspection challenges. • Define model performance KPIs—including precision, recall, F1 score, mAP, cycle-time SLAs, and false-positive budgets—and enforce model governance policies across all GAIO sites. • Lead transfer learning, few-shot learning, and synthetic data-augmentation strategies to reduce reliance on large, manually labeled datasets in new manufacturing contexts. • Integrate industrial camera protocols (GigE Vision, GenICam, RTSP, and ONVIF), and validate camera hardware, exposure and gain settings, and acquisition pipelines on Linux. 2. Edge AI Infrastructure & GPU Optimization • Architect and maintain edge AI deployment infrastructure, including GPU-equipped industrial PCs, NVIDIA Jetson platforms, and containerized inference services on the production floor. • Optimize neural network inference for constrained edge hardware using TensorRT, ONNX Runtime, and OpenVINO, including FP16/INT8 quantization, model pruning, and hardware specific profiling. • Design GPU resource-management strategies for centralized inference servers that support multiple production lines simultaneously, including VRAM allocation, multi-tenancy, and scheduling. • Define, document, and enforce edge-node provisioning standards, including OS image hardening, package manifests, TLS certificate management, and KubeEdge node-registration procedures. 3. MLOps Platform & Lifecycle Governance • Own and operate the GAIO MLOps platform, including experiment tracking, model registry and versioning, CI/CD for model artifacts, staging environments, and automated production rollback procedures. • Implement automated model-retraining triggers based on production data drift, quality-metric degradation, and manufacturing process changes. • Establish data-versioning (DVC, Delta Lake) and feature-store strategies to ensure reproducibility, traceability, and regulatory auditability for production model decisions. • Define and enforce model-inference availability SLAs of 99.9% or higher, establish on-call runbooks, and lead incident response for AI system failures across GAIO sites. • Implement and maintain CI/CD pipelines (GitHub Actions and GitLab CI) for model packaging, edge deployment, and system-health monitoring, including model drift, inference latency, GPU utilization, and node health. 4. Full-Stack AI Application Development • Develop and maintain RESTful APIs, gRPC APIs, and WebSocket services using FastAPI to serve ML models in real time across high-throughput, multi-camera inspection stations. • Build scalable inference pipelines that handle concurrent streams from multiple GigE cameras at production-line throughput.
5. Linux Systems & Container Orchestration • Administer Linux-based edge nodes and central servers (Ubuntu/RHEL) running AI inference workloads, including system-service management, SSH hardening, process monitoring, storage management, and DNS/NTP configuration. • Containerize AI models and full-stack services using Docker, and manage Kubernetes deployments with Helm and KubeEdge to support edge-native orchestration and disconnected plant operations. • Perform systematic Linux-level diagnostics across CPU, memory, disk, and network layers, and produce structured evidence for specialist escalation when issues fall outside the role’s scope. • Manage OS image hardening, patching, security-baseline configuration, and TLS certificate rotation across GAIO edge nodes. • Automate infrastructure provisioning and configuration management using Ansible, Bash, and Python. 6. Industrial AI Rollout & Multi-Site Technical Leadership • Lead end-to-end GAIO site deployments, coordinating with IT, Industrial Engineering, Quality, PLC/Controls, Procurement, and plant management to move AI inspection solutions from approved models to stable production. • Define and own site-readiness gates, RACI matrices, technical risk reviews, acceptance-test protocols, and hypercare periods for each deployment. • Create and maintain reusable deployment runbooks, troubleshooting trees covering device, OS, network, orchestration, application, model, camera, and PLC fault domains, and handover documentation. • Mentor junior AI engineers and site leads in ML/MLOps practices, Linux diagnostics, network troubleshooting, and structured incident-evidence collection. • Serve as the primary technical escalation point between plant sites and central AI/IT teams, translating plant-floor issues into precise, actionable technical tickets for specialist resolution. 7. Security, Identity & Observability • Lead security reviews for GAIO deployments, covering SSO/AD group configuration, TLS/mTLS certificate management, RBAC policies, image-vulnerability scanning, and Microsoft Defender telemetry. • Design and maintain centralized observability stacks (Prometheus, Grafana, Loki, or equivalent) for 20+ GAIO edge nodes across multiple plants, including node health, model-performance metrics, alerting, and runbook integration. • Ensure secure remote-access architecture using VPNs, jump hosts, and zero-trust principles in distributed plant environments with mixed OT/IT network topologies.

Work arrangement
Hybrid