From Preproduction–Production–Postproduction to Layered Assembly Production
by Markus Tischner, Friedrich-Alexander Universität Erlangen-Nürnberg (FAU), Germany.
Many of us in higher education produce under the same pressures. Budgets are tight. Time is limited. Stakeholders care deeply about accuracy and approval safety, but they are often not media experts. At the same time, expectations for engaging, visual storytelling keep rising. Generative AI could change this situation in a very practical way. It could make early prototypes easier, variation cheaper, and “good-enough” assets available faster. Used well, that can help small teams move beyond the default talking head without automatically triggering expensive shooting days or longer schedules.
The question is: Could AI really expand what is feasible in educational media production? If temporary voice, quick visuals, and rapid script iterations are suddenly available, what would be the most useful way to organise the work?
The workflow we know, and the new opportunity
For decades, media production has followed a clear sequence. Preproduction clarifies goals, format, script, and planning. Production creates the material through filming, recording, animation, or design. Postproduction edits, refines, and delivers. This phase model still works in many settings, and it will continue to. It is reliable. It explains roles and costs. It is often the right approach when production steps are expensive and changes must be controlled.
When AI and fast sourcing become part of the process, a new opportunity opens up. Teams can make the media artefact visible much earlier than in a classic phase workflow. Many institutions have not fully tapped into this yet, because AI is often still used as a bolt-on tool rather than something that reshapes how production is organised.
Sidenote: Most higher education teams will not move to “all AI” anyway, and audience feedback often indicates that hybrid results are preferred over fully AI-generated media. What works in practice is hybrid assembly: targeted self-shot material for trust and specificity, carefully selected stock for context, and AI-generated elements for speed, prototyping, and variation.
This hybrid mix fits the realities of educational media production, especially limited budgets and tight timelines. It calls for a workflow that treats hybrid sourcing as normal and stays focused on early assembly and deliberate stabilisation, instead of relying on a single “big production moment” followed by long postproduction.
Layered Assembly Production as a lens, not a finished framework
That is the idea behind Layered Assembly Production (LAP). LAP is not a finished framework to adopt. It is a concept that describes what production can become when assets are easy to generate or source and easy to swap. It is a stimulus to look at daily work differently and adjust parts of the workflow without abandoning what still works.
The contrast is simple. In the classic phase workflow, the timeline becomes relevant late. In LAP, the timeline becomes relevant early. Teams assemble a quick and dirty rough version while the concept and script are still evolving. This early assembly is not meant to look finished. It is meant to be watchable. It lets the team test length, pacing, clarity, and tone while changes are still cheap.
How LAP works in practice: layers that meet in the timeline
To make parallel work manageable, LAP describes five layers that need to stay aligned. The point is not to create five departments. The point is to keep the essential types of work connected while the timeline evolves.
Story covers learning goals, structure, script, and claims. In LAP, story is refined while timing is already visible, which makes script decisions more realistic.
Visuals include own shots, stock, AI visuals, graphics, and style rules. Visuals start as placeholders and become more specific in versions. Consistency needs active attention.
Audio covers voice, music, ambience, and effects. Temporary voice is a timing tool, and voice choices are often trust choices in educational contexts.
Timeline is the integration space where pacing and meaning can be judged as a whole.
Integrity includes accuracy, implied meaning, rights, transparency, and accessibility. In LAP, integrity work cannot be postponed, because prototypes already imply claims and mixed sourcing increases complexity.
The rough assembly is where these layers meet. Even a simple first cut already behaves like a film. That is why it changes decision-making.
A specific advantage for CTL–educator collaboration
A typical challenge in higher education is co-creating with educators who are not used to cinematic constraints. Many underestimate how short a film text must be compared to written formats. They may struggle to evaluate pacing, rhythm, and visual meaning based on a script alone. In the traditional phase workflow, the film becomes visible late, which can lead to long revision loops and late text expansions.
LAP improves this collaboration because it makes the film visible early. Educators can react to a prototype instead of imagining the outcome. Feedback becomes more film-relevant, because people can comment on density, clarity, and emphasis as experienced in time. Approvals become clearer because decisions happen on the viewing experience, not on abstract drafts. For CTL teams, this reduces translation work and stabilises expectations earlier.
Locking points: how to avoid endless iteration
AI makes alternatives cheap. That is helpful for exploration, but it creates a real risk: endless iteration. When every change is easy, nothing feels final. Teams keep generating “one more option,” and the process loses direction.
LAP counters this with deliberate locking points. A locking point is a decision moment where the team agrees: from now on, changes require a reason and a tradeoff. Locking points are not meant to block creativity. They protect progress.
Typical locks include:
Script lock: meaning is stable and duration is within range. Wording can still be refined, but new ideas stop entering the script.
Anchor lock: the grounding strategy for visuals is stable, so later choices align with the same world.
Voice lock: the trust strategy for voice is stable and validated where possible.
Integrity lock: rights, transparency, and accessibility obligations are covered and documented for release.
Join the discussion in Leuven
This topic will be featured as a Discussion Session at the next Media and Learning Conference in Leuven, 17–18 June 2026, with a short input and a panel discussion with experienced practitioners. If you produce media in higher education, you will recognise parts of this shift. The useful question is no longer only “Which AI tool should we use.” The more valuable question is “Which workflow changes help us use AI and hybrid sourcing responsibly, efficiently, and with consistent quality.”
If LAP does its job, it offers a clearer lens on what is becoming possible and a practical nudge to adapt parts of your workflow, without abandoning what still works.
Markus Tischner, Friedrich-Alexander Universität Erlangen-Nürnberg (FAU), Germany
