Demonstrators for Industrial Cyber-Physical System Research: A Requirements Hierarchy Driven by Software-Intensive Design
This paper proposes a five-level hierarchical framework for defining demonstrator requirements in software-intensive industrial cyber-physical systems to address the common mismatch between project goals and achievable results caused by vague elicitation practices.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine you are leading a massive, multi-year cooking competition. You have a team of chefs (researchers), a list of ingredients (data and code), and a goal to create a "Grand Feast" (the research demonstrator) to show off at the end.
The problem, according to this paper, is that when the competition starts, everyone has a different idea of what the "Grand Feast" actually looks like. Some chefs think they just need to prove they can chop an onion (a basic proof). Others think they need to serve a fully plated, Michelin-star meal to a real crowd (an industrial-grade system).
Because no one agreed on the menu or the definition of "ready" at the very beginning, the team spends years arguing, missing deadlines, or serving a half-baked dish that no one is happy with.
Here is how the authors of this paper propose to fix that chaos.
The Problem: The "Magic Wand" of TRLs
In the world of research, people often use a ruler called the TRL Scale (Technology Readiness Level) to measure how "ready" a technology is. It goes from Level 1 (a vague idea) to Level 9 (a fully working product).
The authors say this ruler is like trying to measure a cake's taste with a ruler. It tells you the size, but not if the cake is actually edible or if the ingredients match.
- The Issue: A project might say, "We will build a Level 6 demonstrator!" But they haven't defined what that means. Is it one chef cooking alone? Is it five chefs working together? Does it need to taste good to a real customer, or just look good in a photo?
- The Result: Confusion. The academic chefs want to show off a cool new recipe; the industrial partners want a machine that works in a real factory. They end up with mismatched expectations.
The Solution: A New "Menu" (The Taxonomy)
The authors created a new, more detailed menu with 5 specific levels of "demonstrators." Instead of just saying "Level 6," they ask:
- Who is cooking? Is it just one chef (one part of the project), a few chefs working together, or the whole kitchen?
- What are they serving? Are they just showing the food exists (Functional), or are they showing it's fast, reliable, and tasty (Extra-functional)?
- Who is eating? Is it a single customer (one use-case), or a coordinated group of customers?
They call these levels things like "Proof of Concept" (just showing it works) and "Optimised Grand Proof of Integration" (the whole team serving a perfect meal to a real crowd). This helps everyone agree on exactly what the final dish should look like before they start chopping.
The Tool: The "Pre-Game Checklist" (The Framework)
To make sure the team doesn't get stuck halfway through, the authors built a 7-step checklist (a framework) to use before the project even starts.
Think of this as a "Reality Check" meeting before the cooking begins. You take three things:
- The Proposal: What did we promise to do?
- The Plan: How do the chefs depend on each other? (e.g., Chef A can't start until Chef B finishes the sauce).
- The Ingredients: Do we actually have the raw data and code from the industrial partners?
The checklist runs through these steps:
- Step 1-3: It looks at the plan and asks, "If Chef A is slow, does that stop Chef B?" It finds the weak links in the chain.
- Step 4-5: It checks the ingredients. "Do we actually have the data from the factory, or is it just a promise?"
- Step 6: It matches the reality to the menu. "Okay, we promised a Level 6 feast, but we only have enough ingredients for a Level 3 potluck. Let's adjust the menu now, not later."
- Step 7: It writes down the new, realistic rules for the team.
Real-World Examples (The Test Kitchen)
The authors tested this checklist on two real research projects:
1. The ZORRO Project (Early Stage)
- The Situation: The team promised to build a "Level 6" industrial machine.
- The Check: The checklist looked at the dependencies and realized the "ingredients" (data from specific companies) didn't match the "chefs" (the software teams). The companies providing data weren't connected to the teams that needed them.
- The Fix: The framework told them, "You can't build the Level 6 machine yet. You have to either change the plan to connect the right people, or lower your goal to a smaller, simpler demo." This saved them from building something impossible.
2. The PrimaVera Project (Late Stage)
- The Situation: This project was almost finished. They had promised a "Digital Twin" (a perfect virtual copy of a ship) that integrated everything.
- The Check: Looking back, the authors saw that the team had to scramble to find the right partners because the initial plan didn't match the reality of who had the data. They ended up with a "product catalogue" of small demos instead of the one big "Grand Feast" they promised.
- The Lesson: If they had used the checklist at the start, they would have seen the mismatch immediately and adjusted their goals to be realistic, avoiding the last-minute scramble.
The Bottom Line
The paper argues that research projects often fail to deliver what they promise because they treat "demonstrators" as a vague idea rather than a specific, measurable goal.
By using this new 5-level menu and the 7-step checklist, research teams can stop guessing. They can look at their ingredients and their team structure, realize what is actually possible, and set a goal that is ambitious but achievable. It's about making sure the chefs and the customers agree on the menu before the first pot is even heated.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.