AI Vision and Automation Engineer (KTP Associate) - Full Time, Fixed Term up to 24 months duration

This post is funded through a Knowledge Transfer Partnership (KTP) between Harper Adams University and Karsten UK Limited, a fresh citrus processing and distribution business supplying major UK and European retailers. The Associate will develop and embed an AI-driven vision and quality sorting capability at Karsten UK's Newport facility. A Fanuc Delta pick-and-place robot, already owned by Karsten UK, provides the actuation platform. The system will inspect and grade fruit against retailer specifications, using a compact conveyor training rig. The KTP funds the vision and training components, not the robot.

Working from the conveyor training rig, the Associate will develop a Discrete Element Method (DEM) simulation model of fruit dynamics. The Associate will build a photorealistic rendering pipeline to generate synthetic training images with automated annotation, and train AI classification models to achieve 95% accuracy using synthetic data alone. Integration with the robot controller will then translate classification outputs into pick-and-place sorting. The training method is product-agnostic and can be retrained for new products and specifications.

The role combines AI model development, simulation-driven data generation, robotic integration, and practical engineering delivery in a food processing environment. Employment will be full-time at Karsten UK's Newport facility, directed day to day by the site team. The Associate will receive at least half a day per week of academic supervision from Harper Adams University, with access to its Collaborative Simulation Laboratory. Development, training, and validation will take place on the conveyor training rig beside production, and live production will be reached only after validation on the rig.

The appointee will also be expected to:

Develop and deliver training programmes for production and maintenance personnel to embed the vision and sorting capability within Karsten UK's operational teams. Establish performance monitoring dashboards and continuous improvement protocols to sustain system performance. Quantify and report commercial benefits, including throughput improvement and value recovery from more accurate grading. Prepare technical reports, academic publications, and a project case study in accordance with KTP programme requirements, and present progress at regular Local Management Committee meetings involving academic and industrial supervisors. Comply with the health and safety, food safety, data management, and intellectual property policies of both Harper Adams University and Karsten UK Limited.

Essential Requirements:

A first or upper second-class honours degree in Computer Science, Electronic Engineering, Mechanical Engineering, Mechatronics, Robotics, or a closely related discipline. Demonstrable experience developing and training AI or machine learning models for image classification or object detection. Proven programming or model-development experience in Python and/or MATLAB & Simulink, including scripting for data processing, model training, and automation workflows. Understanding of computer vision principles including image formation, lighting, and camera calibration. Strong written and verbal communication skills. Ability to manage complex technical work independently and within a team.

Desirable Requirements:

A postgraduate qualification (MSc or PhD) in a relevant engineering, computer science, robotics, or automation discipline. Experience with physics-based simulation tools such as DEM, FEA, CFD, or multibody dynamics software. Familiarity with photorealistic rendering software such as Blender or KeyShot for synthetic data generation. Knowledge of industrial control systems, PLC programming, or robot programming. Practical experience integrating vision or machine learning outputs with robotic, automation, or mechatronic hardware. A background in food processing, agricultural, or manufacturing environments. Familiarity with version control systems and collaborative software development practices. Prior industrial placement or collaborative research demonstrating delivery of practical outcomes.