How Celean’s open data approach advances lean engineering knowledge

Lean engineering turns principles such as pull, flow, quality at source and continuous improvement into capabilities that can be designed, measured and improved. The hard part is turning work on a shop floor, in a project team or across a supply chain into knowledge that other people can understand and test. A task completed once is useful; a lesson carried into the next design cycle is valuable.

Celean Team, the Chalmers Electronic Lean Research Team, works across Chalmers University and industry to study digital visual planning, Kanban, visual project management, enterprise systems and lean processes. Its open data approach adds a crucial layer: research outputs such as methods, datasets, models, prototypes and field observations can become inspectable evidence rather than isolated private artefacts. The aim extends beyond publishing files, retaining enough context for others to judge and reuse them.

This matters because much operational expertise remains tacit. Experienced operators know why a queue keeps forming, while engineers know which constraint changed; neither insight becomes useful across an organisation if it is held only in memory. Open, well-described evidence gives teams a shared basis for discussion and comparison. In Australia, where engineering and operations span metropolitan hubs, regional suppliers and remote sites, that common evidence base can be especially valuable.

Turning operational signals into engineering knowledge

The Project Visit website brings together research publications, project information, prototype software tools and examples of collaboration with industry. For Celean, these are complementary routes into the same question: how can digital tools support lean behaviour in real work? The open data approach treats a publication, a project update and a working prototype as complementary rather than competing outputs. Each reveals a different layer of the problem, from an abstract method to the behaviour it produces in a team.

Digital visual planning makes sequences of work, work-in-progress limits, blocked cards and planned flow visible on a shared board. Kanban adds a pull logic: downstream needs determine when upstream work may begin. When those boards connect with enterprise systems, each visual signal can be paired with data about capacity, quality, delivery or cost. This combination turns “the board looks busy” into a more precise question: where is flow being lost, and what evidence explains it?

Over repeated cycles, those records become engineering knowledge. Standard work, takt, staffing and exception handling can be compared before and after a change. A result that remains only in one team’s memory cannot challenge a design assumption elsewhere. Shared data allows other teams to see the pattern, test it in a different context and report exceptions. Knowledge therefore grows through an evidence loop rather than through authority or anecdote alone. The value is cumulative: one observation rarely settles a design question, but comparable records can show whether the same mechanism appears across shifts, products or projects.

What makes research evidence genuinely reusable

Open data becomes useful only when another person can interpret it without the original researcher standing beside them. Metadata and provenance are not administrative extras; they preserve the assumptions that make a result intelligible. A strong evidence package for lean engineering work can include:

• A clear question and intended use, so users know which operational decision the evidence can support.
• Consistent definitions for flow, lead time, work in progress, demand and performance measures across datasets.
• Version history that shows when methods, models or interface rules changed and why.
• Traceable provenance linking summaries and visualisations to their source records and processing steps.
• Stated limitations, including context, missing data, sampling effects and situations where findings should not be applied.

These details make comparison safer and help prevent a promising prototype from becoming a misleading general rule. For teams working with the resulting knowledge base, the practical effects include:

• Faster onboarding, because new engineers and operators can study real cases instead of relying only on instructions.
• Better root-cause analysis, as recurring delay, quality or staffing patterns can be traced across related work.
• Stronger standard work, since proposed changes can be connected to observed performance and revision history.
• More comparable experiments, allowing pilot teams to use common measures while recording local conditions.
• Auditable decisions, making it easier to explain why a workflow, resource allocation or policy changed.

The social effect matters as much as the technical one. Shared records give designers, suppliers, operators and managers permission to point to the same evidence. They also make disagreement productive: teams can debate which explanation fits the data rather than whether a personal account sounds convincing. This is how tacit knowledge becomes partly explicit without pretending that a dataset captures every practical skill. It also supports a more honest account of improvement, because evidence from different sites can reveal where a solution transfers and where local conditions require adaptation.

From Kanban boards to connected enterprise systems

Kanban is often described as a simple board, but its real value lies in the rules represented by that board: visible work, explicit priorities, limited work in progress and controlled pull. Digital implementation expands what teams can see without changing those principles. A card may move across a value stream while status, queue time, blocked time and ownership remain attached to it. Enterprise systems can supply the surrounding context, making it possible to relate daily behaviour to longer lead times. That distinction prevents digital sophistication from being mistaken for operational rigour.

The visual plan becomes a shared interface between planning and operations. Instead of a weekly progress meeting built on conflicting spreadsheets, teams can compare intended sequence with actual flow. When the same definitions govern procurement, production, maintenance and project delivery, local improvements reveal their effect on the whole system. This interoperability turns visual management into more than a reporting routine. Open data also allows selected, appropriately protected evidence to cross organisational boundaries, which is valuable when one company’s bottleneck simply moves to a supplier.

A dashboard alone can create false confidence. If measures change, timestamps are missing or operational context is stripped away, a polished chart may hide the very problem it appears to solve. A sound open approach therefore preserves workflow states, assumptions, rejected alternatives and negative results. It links the visual record back to its origin and treats data quality as part of the method. In this setting, visualisation is not decoration; it is a way of exposing where a lean design meets reality.

Collaboration that travels across industry and geography

Celean’s position between university research and industrial practice gives its work a route from formal study to operating conditions. The wider collaboration includes organisations such as Volvo, where questions about visual management, production flow and digital support can be examined against substantial engineering environments. Chalmers contributes research methods and conceptual rigour, while industry contributes complexity, constraints and experience. The exchange works best when evidence returns in usable form and practical findings become topics for further study.

Australian application has its own geography and market structure. Melbourne combines advanced manufacturing, infrastructure and logistics; Brisbane links transport, construction and regional supply chains; and Pilbara operations depend on dispersed sites, long travel distances and demanding handover conditions. Large multinational operations work alongside small and medium-sized suppliers, engineering firms and specialist contractors. Research must therefore survive different enterprise systems, purchasing rules and levels of digital maturity. A method that works only inside one large company’s integrated platform is not yet robust open knowledge. In that market, a reusable method must be easy to explain, comparatively easy to pilot, and capable of producing evidence that survives a handover between organisations.

Communication is part of that robustness. Australian teams commonly favour a direct, fair-dinkum update that exposes the issue without assigning blame, and a short team retrospective may be called a wrap. Plain language is valuable here because it lets an operator, a software developer and a procurement specialist discuss the same constraint. The research and project updates illustrate how progress can be communicated around a concrete problem, a prototype or a lesson from implementation. Such updates also give cross-border and cross-site partners a way to maintain context when they cannot share a physical board.

Learning in public, one improvement cycle at a time

Celean Team’s open data approach advances lean engineering knowledge through a disciplined plan-do-study-act loop: observe work, represent it visually, share the evidence, test a change and return the result to the community of practice. Publishing at every stage makes the process inspectable. It allows others to distinguish a sound principle from an artefact that worked only under favourable conditions, and it makes neglected cases, such as failed pilots or social resistance, part of the record rather than footnotes.

Open does not mean careless. Intellectual property, privacy, commercial sensitivity and safety-critical information require proportionate controls, including aggregation, access rules or synthetic examples. The purpose of governance is to protect people and legitimate interests while preserving the traceability that makes learning credible. Data must also be checked for completeness, consistency and representativeness. Poorly governed openness can spread uncertainty faster than good practice, just as an unmanaged metric can turn improvement into theatre. Clear ownership and review dates keep this material current, while transparent decisions about what may be shared prevent openness from becoming an accidental disclosure.

The lesson to remember is that a Kanban card, visual plan, prototype, dataset and field note become more valuable when they preserve context, invite comparison and feed the next improvement cycle. Celean’s contribution is to connect those artefacts through open, responsible research, turning scattered operational experience into cumulative lean engineering knowledge that teams can test, extend and refine.