PyTorchPythonTensorFlowUnity
業務内容
- Design, implement, and enhance ML-based driver workload estimation algorithms using vehicle CAN signals, in-vehicle sensors, and driver operation logs as time-series inputs
- Build and operate training data infrastructure and evaluation pipelines, including label/annotation policy design, data preprocessing, and feature engineering
- Investigate and improve logic that combines CAN-based workload indicators with other measures (surrounding-vehicle and vision-based indicators)
- Design interfaces to feed workload estimation results into driving-suggestion and voice-prompt logic; continuously improve overall feature performance
- Collaborate with software, test, UX, and other stakeholders to align requirements and evaluation metrics, and share experiment results and learnings
- Participate in planning and conducting evaluations using simulators and on-road test vehicles, driving improvement cycles from real-world findings
技術スタック
必須スキル
- 3+ years of practical experience in software/algorithm development using vehicle CAN signals, in-vehicle sensors, and driver operation logs as time-series data
- Practical experience with machine learning, deep learning, and/or statistical modeling
- Python development with ML/DL frameworks (e.g., PyTorch, TensorFlow)
- Ability to independently drive end-to-end ML development from data preprocessing and feature design through training and evaluation
- Strong communication skills to work with multiple stakeholders
- Willingness to travel for business purposes
- Business-level Japanese and conversational English
歓迎スキル（該当する場合）
- Ability and willingness to work at the Susono office
- Experience in in-vehicle systems, ADAS, driver monitoring, or robotics
- Experience in feature engineering and developing driver workload/state estimation using CAN signals, sensors, and time-series data
- Experience deploying ML models to production (edge/embedded optimization, inference pipeline, MLOps)
- Experience using simulators (e.g., Unity) for evaluation or data generation
- External outputs in ML or signal processing (competitions, publications)
- Bachelor’s degree in CS, electrical/electrical engineering, control engineering, or related field, or equivalent practical experience
キャリア成長観点
- 専門性の深化: ドライバー workload推定、センサ融合、時系列データを扱うMLの高度化をリードできる
- 影響力の大きさ: 車載AI・ADAS領域の安全性・快適性向上に直接寄与する機能開発に携われる
- 複合スキルの習得: ソフトウェア、評価、UX、シミュレータ、実車評価を横断してプロダクト価値へ落とす経験を積める
- 実運用への展開: edge/組み込みデプロイ、Inferenceパイプライン、モデル運用などの実装経験を獲得
- キャリアパスの幅: MLエンジニアとしての高度化だけでなく、AIシステムエンジニア、データサイエンティスト、ロボティクス/自動運転周辺領域への展開が見込める
企業についての所感
高年収かつ高難易度の選考プロセスの企業。車両OSのArene、スマートシティのWoven Cityを支えるプラットフォームの開発など規模感の大きい開発をしている。
データ取得日: 2026/9/10