IDEA

Blog

Our perspective on engineering services.

Practical examples from our work and our views on the methods and tools used to create products.

Comparison of two learning curves: starting from zero versus IDEA. Thanks to cross-disciplinary knowledge, IDEA starts at an elevated level, picks up new knowledge more steeply, and is ready for deployment considerably sooner – the time advantage is marked.CompetenceTime →ready for usefrom scratchIDEAStart: Broad know-howTime advantage
Starting from scratch takes a long time to reach operational capability in a new field. Existing cross-sectional knowledge raises the starting point and steepens the curve – your input know-how (NDA) accelerates this further. Schematic.
95
Perspective · 2 min

Learning fast is the business

New products emerge where there is no manual. Why quickly getting to grips with knowledge is our core skill – and how your know-how accelerates it.

Read →
Requirements hierarchy from top to bottom: the most important requirement at the top, subordinate ones below.Requirements hierarchy1Look and feel must match the company2Climate in the room must be perfect3Costs must not be exceeded4Innovative lifting system for the USP5Versatile design· · ·and many moremore important
Requirements hierarchy from top to bottom: the most important requirement at the top. A real project has many more – if two contradict each other, the higher one wins.
94
Perspective · 2 min

Everything begins with the customer. With your customer!

Why every development at our company starts with the customer, and a hierarchically ordered list of requirements anticipates decisions.

Read →
The product development process from idea to series production, with three examples below showing how IDEA (blue) and your team (green) share the stages.IdeaConceptDevelopmentPrototypeManufacturingSeriesProduct development →Everything from a single sourceAdd selectivelyOnly developmentIDEAYour team
You can take this path entirely with us, split the stages – or bring us in for development only. Blue is what we handle, green is your team. You decide where the line is drawn.
24
Perspective · 2 min

From One-Stop Shop to Individual Item

The whole path from idea to series – or just the parts you're missing.

Read →
Generic project dashboard: section switcher, combinable filters and an artefact matrix with version, manufacturing stages, status and feedback per variant.PROJECT DASHBOARDon trackConceptCADPrototypeManufacturingTestingVariantsFeedbackFilterAllIn progressIn testingApprovedDiscardedPreviewDateVariantVer.CADExportProductionStatusReturn12.06.Housing Av4In testing3 / 410.06.Bracketv2In progress5 / 509.06.Cover Bv1In progress–07.06.Adapterv3In testing2 / 305.06.Framev6In progress4 / 403.06.Flapv2In progress1 / 2
This is how complex tracking can become: dozens of variants, each with its own version, manufacturing stage, status and returns. A good dashboard filters and condenses this – until you can see at a glance where the project stands. Schematic, values anonymised.
05
Perspective · 2 min

Communication is key

Why communication determines project success – and how we become your project memory.

Read →
A requirements specification divides the overall system into individual trades, each in turn into clearly named work packages.Overall systemFoundationBase plateAnchoringLevellingDamping· · ·Trade 1Support frameBracingWelded partFittings· · ·ElectronicsControl systemWiringSensorsControl cabinet· · ·Trade 2DriveGearboxBearingCoupling· · ·TrailerChassisDrawbarBrakeLighting· · ·Trade 3CladdingSealsPaintingAssembly· · ·
The overall project is divided into individual trades, and each individual trade in turn into its work packages. This divisibility forms the basis for obtaining quotations and awarding contracts.
60
Method · 2 min

What a requirements specification delivers — and when you need one

How a requirements specification makes a complex product divisible – separable units, comparable offers, a shared reference.

Read →
Two areas: What AI takes over – variants, research, documentation drafts, routine work – versus what the engineer is responsible for: physics and safety, liability, trade-offs, and judgement.What AI takes overWhat the engineer ownsGenerate variantsResearch & review standardsDocumentation draftsRoutine workPhysics & SafetyLiability & ApprovalDeciding trade-offsContext & Judgement
The left column can be delegated. The right one cannot – it is about responsibility, and that is carried by a human being.
35
Perspective · 2 min

The Opportunity and Limits of AI in Product Development

Where AI genuinely helps in development – and where the engineer remains responsible.

Read →
The AI produces a first draft, an engineer checks it technically and against the standards, and the result is robust documentation.AI first draftfrom design dataEngineer reviewstechnical · compliantReliable documentationverified, up to date
The order is reversed: the AI delivers the draft in hours, the engineer checks it technically and against the relevant standards. The documentation is the responsibility of the human – and stays up to date, because a new draft costs next to nothing.
75
Method · 3 min

Technical documentation with AI

Why documentation goes out of date – and what changes when the first draft only takes hours.

Read →
The reverse engineering process as a chain: disassemble, measure and classify, build a parametric CAD model, and finally drawings and bill of materials.DisassemblingAssembly & joining1Measuring & classifyingTolerance key dimensions2CAD modelparametric3Drawing setwith bill of materials (BOM)4
Each step builds on the previous one. The green dots are the transitions – only once a step is solid does its result serve as a valid input for the next.
87
Method · 3 min

Reverse Engineering: The Part Exists, the Supplier Doesn't

A part that no one can manufacture any more – and how it becomes a design again.

Read →
The cost of a change increases steeply over the course of a project. Decided early, changes are cheap; decided late, they are expensive.Feasibility studysmall budgethigh riskCost of a changeConceptDevelopmentImplementationOperation
At the start, the design freedom is large and change is cheap. With every step, the cost of a correction rises – not linearly, but steeply. That's why the leverage lies at the front.
34
Method · 2 min

Feasibility study: when it pays off

A small budget for clarity, before a large one flows blindly into development.

Read →
The seven steps of CE conformity as a sequential process: from the applicable legal acts to the affixed CE mark.1Determine applicable directives and regulations2Carry out risk assessment3Identify applicable harmonised standards4Prepare technical documentation5Complete conformity assessment procedure6Issue EU declaration of conformity7Affix CE marking
The order is fixed, because each step depends on the results of the previous one; the CE marking is therefore applied last.
02
Compliance · 3 min

CE marking: an overview of the steps

From the directive to the CE mark – conformity in seven steps.

Read →
Our offer stays within the agreed min–max corridor. The competitor's lowball offer starts below it and later breaks through it on the upside.CostProject progress →Our corridor · guaranteed Min–MaxLoss leaderactual costs
The lure offer starts below our corridor – and ends up well above it. Our offer stays within the agreed band: never below the minimum, never above the maximum value.
22
Costs · 2 min

What Product Development Really Costs – A Realistic View

Time and materials billing – with a guaranteed min-max corridor for the agreed scope. And why cheap bait-and-switch offers end up costing more.

Read →
KPI development over time: flat until the introduction of the innovation, then rising. The gain is shared between the customer and IDEA.KPITime →Baseline valueInnovation introducedYour profitIDEA fee
Before the innovation, the KPI runs flat. Afterwards, it rises – and only the gain above that line is shared: half remains your profit, the other half is our fee.
40
Costs · 2 min

Innovation in Engineering Billing

A fee tied to success: we receive a pre-agreed share of the measured gain.

Read →
Two approaches compared. Top: the usual approach: the tool depends on a cloud AI service during operation, so company data leaves the premises with every use. Bottom: the IDEA approach: AI builds the tool once, and the finished tool then keeps running in the company without internet or an AI service.USUAL PATHYOUR COMPANYYour dataToolAI service · Cloudwith every useSubscription · DependencyWITH IDEAYOUR COMPANYYour dataToolAI builds the toolonce, at build timeafterwards no internet
If AI is embedded in ongoing operation, something leaves the building with every use. If it only builds the tool, everything stays within the company afterwards.
69
Perspective · 4 min

AI and Data Protection: Tools That Stay Offline

Fear of data leaks through AI tools is justified. Why our tools still work without internet and without an AI service.

Read →
A product at the centre, from which five axes radiate — one each for function, design and manufacturability, appearance, anticipated manufacturing costs, and risks regarding standards and conformity.The productFunctionDoes it do what it was acquired for?· Core benefit· Edge cases· RepeatabilityDesignIs it built solidly and manufacturably?· Material· Tolerances· Joining technique· MaintenanceAppearanceHow does it feel on first contact?· Operation· Haptics· Surface· Joint gapsManufacturing costsWhat should it cost, in principle?· Machining· Number of parts· Should-costStandardsWhich rules apply, where is action needed?· CE· harmonised standards· marking
Five axes, one report: function, design, appearance, manufacturing costs and standards risks are assessed separately and put into context with one another.
18
Method · 3 min

Product Audit – An Outside Perspective on Your Own Product

Anyone who looks after a product for years eventually stops seeing what a buyer notices on first contact. An audit brings a fresh engineering perspective to it.

Read →
Three curves of cumulative costs over 60 months. Pure outsourcing rises linearly in IDEA blue. In-house hiring follows the outsourcing curve until month 6, runs steeply during the training phase and flattens from month 12; an orange band above it shows the risk premium. The outsourcing curve crosses the hiring curve in the expected case at month 13, in the worst case at month 20.0200400600800Cumulative costs (thousand €)01224364860MonthsPosition filledTraining period endsBreak-even ≈ month 13Break-even including risk ≈ month 20Purchasing (contract for work)Employment — Expectation+ Risk (bad hire, parental leave, sick leave)
Cumulative costs over 60 months. Buying in rises linearly from the first month (blue). Hiring starts after a vacancy search, works at half productivity during the onboarding phase, and carries its own weight thereafter (orange, filled). The hatched band shows the risk premium from a bad hire, long-term sick leave, parental leave, and onboarding delays. The break-even point is around month 13 in the expected case, or month 20 at the earliest with the risk premium applied.
31
Costs · 5 min

Hiring or outsourcing – when responsibility comes at a cost

An employed engineer is not liable for their design, but an engineering office is. This shift in liability comes at a price that a simple hourly-rate calculation does not capture.

Read →
27
Standards & conformity · 8 min

Machinery Regulation: What changes on 20 January 2027

On 20 January 2027, the Machinery Regulation replaces the Machinery Directive. Every unit delivered from that day on must meet new requirements, above all with regard to software and cybersecurity.

Read →

Are you thinking of a specific project?

The more precisely you describe what you need, the more precisely we can tell you what it costs.

Request a project
Blog – AI, Engineering & Product Development | IDEA