Ideation Support Agent
Project description
The primary challenge of this project was to enhance the capture, retrieval, and reuse of ideas, problems, and solutions within Vanderlande. Employees often struggle to find relevant past ideas, leading to duplicated work and lost knowledge.
The project explored how an AI-powered ideation assistant could support employees by retrieving existing ideas, suggesting related solutions, and guiding users through structured ideation. The outcome is a Copilot-based agent integrated with SharePoint that supports ideation, knowledge reuse, and decision-making, while aligning with organizational constraints such as data governance and security.
Context
This project was executed within the Innovation team at Vanderlande, focusing on digital innovation and continuous improvement in the industrial automation and logistics domain. The Innovation team supports employees across the organization in developing, validating, and scaling new ideas. However, innovation-related knowledge is often spread across multiple tools and repositories, making reuse challenging.
The project addresses this challenge by applying AI in a controlled and responsible way to support innovation workflows. The solution aligns with Vanderlande’s broader digital transformation strategy, aiming to enhance collaboration, improve knowledge accessibility, and reduce inefficiencies without disrupting existing processes or systems.
Results
The most important outcome of this project is a working prototype of an AI-powered Ideation Copilot integrated with Vanderlande’s SharePoint environment. The solution enables users to search for existing ideas, retrieve relevant background information, and receive structured guidance during the ideation process. This directly addresses the problem of knowledge fragmentation and reduces the risk of duplicated ideas or missed insights.
Validation was conducted through live user testing sessions with employees from different roles. Users interacted with the agent using real innovation-related questions and provided qualitative feedback on usability, relevance, and clarity of responses. The results showed that the agent successfully supported idea exploration and improved users’ ability to quickly understand whether similar ideas already existed. Feedback also highlighted the value of having a single, conversational access point to innovation knowledge.
From a Technology Readiness Level (TRL) perspective, the solution is positioned at TRL 5–6: a validated prototype tested in a relevant organizational environment. Core functionalities are operational, and technical feasibility has been demonstrated, while further work is required for large-scale deployment, performance optimization, and integration with additional data sources (e.g. other internal tools).
Overall, the project demonstrates clear value for the Innovation team by improving knowledge reuse, supporting better-informed ideation, and providing a scalable foundation for future AI-driven innovation support within Vanderlande.
About the project group
Our project group consisted of ICT & Business students with backgrounds in data analytics, business process optimization, and applied AI. The project was conducted over one semester, with an average workload of 20–24 hours per week. We worked in an iterative, agile manner with regular stakeholder check-ins at Vanderlande, combining desk research, user interviews, and rapid prototyping. Collaboration was done through structured sprint planning, feedback sessions, and hands-on testing with end users. The project focused on delivering a practical, business-ready solution rather than a theoretical concept.
