Know document status
Frameworks, standards, guidelines and laws are not interchangeable.
Reference documents, conformity constraints, V&V, resources and pricing models turn a technical idea into a proposal that can actually be negotiated.
Frameworks, standards, guidelines and laws are not interchangeable.
Applicable EU legislation drives conformity; harmonised standards are usually voluntary routes.
Commercial model changes exposure, but legal/liability allocation remains contractual.
Workstreams, resources, materials, lead times and contingency must explain the number.
A credible Proposal links the technical solution to reference basis, verification evidence, resources, schedule, uncertainty and a commercial model.
The complete English academic notes follow the Ukrainian source, retaining the original argument, engineering examples, terminology, diagrams, tasks and editorial qualifications. The original transcripts remain available separately in Ukrainian.
Chronology: 16 September 2026 · fifth session. Primary source: the complete transcript. Focus: turn a pre-sales vision into a controlled collection of engineering assumptions, regulatory references, resources and commercial risks.
The session begins by revisiting the sequence used as the course's teaching model:
Lead
↓
RFI
↓
RFC (Request for Clarification — course terminology)
↓
Proposal
↘
RFQ / commercial estimate
In real procurement, the sequence may differ; some deliverables may run in parallel or have different names. What matters is not bureaucratic ordering, but the logic: understand the task first, then articulate an approach, and then estimate resources and risks transparently.
For systems interacting with the physical world, customer expectations often cover much more than functions:
A proposal therefore needs a reference basis: the regulatory, process and engineering documents affecting the solution.
The lecture proposes a helpful teaching hierarchy:
PROCESS / ASSESSMENT FRAMEWORK
↓
STANDARDS
↓
GUIDELINES / RULE SETS / PRACTICES
↓
PROJECT IMPLEMENTATION
This must not be mistaken for a universal legal hierarchy, since particular documents can have different legal or contractual status. It remains a useful way of organizing technical reasoning.
Mostly answers the questions “Which processes should exist, and how is their capability assessed?”
Example: Automotive SPICE.
Defines requirements, principles, work products or criteria for a particular domain.
Helps translate high-level requirements into concrete engineering rules. Examples include MISRA C/C++, MAAB style guidelines and internal project rules, but the reasons for applying them must always be stated.
The transcript mentioned a model with seven levels. The capability scale in Automotive SPICE PAM 4.0 actually runs from Level 0 through Level 5:
| Level | Name |
|---|---|
| 0 | Incomplete |
| 1 | Performed |
| 2 | Managed |
| 3 | Established |
| 4 | Predictable |
| 5 | Innovating |
Official VDA QMC PAM 4.0: https://vda-qmc.de/wp-content/uploads/2023/12/Automotive-SPICE-PAM-v40.pdf
It is also important that Automotive SPICE is a process assessment model, not a certificate that automatically authorizes the holder to develop systems of a particular complexity class.
In the spoken lecture, standards were identified too strongly with laws and CE marking with “ISO certification”.
For the European Union, the more accurate model is:
applicable EU legislation = mandatory legal requirements
↓
harmonised standards = usually voluntary route to presumption of conformity
↓
conformity assessment + technical documentation + declaration
↓
CE marking, where required by applicable law
In other words:
Official EU resources:
HARA and ASIL reappear in the lecture as an example of risk classification determining the rigor of engineering activities.
Under ISO 26262, automotive ASIL is classified using Severity, Exposure and Controllability. QM and ASIL A–D must not be treated as a universal scale that applies without adaptation to every industry.
ISO 26262-3: https://www.iso.org/standard/68385.html
Similarly, rules such as “prohibit dynamic memory” or “keep 70% of memory unused” may be project-specific or domain-specific guidelines, but should not be presented as universally explicit ISO 26262 requirements without a precise reference.
The customer must be able to see not only what we will build but how we will demonstrate that the result satisfies the need.
Do not mechanically insert every testing technique into each proposal. The V&V strategy must derive from product risks, requirements and the applicable process or reference basis.
An embedded-project estimate includes at least four resource layers:
| Layer | Examples |
|---|---|
| People | architect, firmware, hardware, mechanics, QA, safety and security, manager |
| Materials | components, PCBs, mechanical parts, harnesses, fixtures |
| External services | fabrication, assembly, laboratories, certification or conformity-assessment work |
| Time / dependencies | lead times, shipment, customs, redesign cycles and waiting for samples |
Consequently, the Gantt chart and critical path of an embedded project frequently depend on things that cannot be accelerated merely by adding another software engineer.
The lecture compares two basic commercial models.
The customer pays for actual engagement at agreed rates and, as provided by the contract, material and external-service expenses. This model works well with uncertain scope, discovery, research, R&D and frequent changes.
The provider commits to completing a defined scope under agreed price and contractual conditions. This requires much clearer scope, acceptance criteria, assumptions and change control.
It is incorrect to claim categorically that “all risks are the customer's under T&M” or “all risks are the vendor's under Fixed Price”. Legal, safety, IP, liability, warranty, schedule and commercial risks are allocated by the contract.
A simplified comparison:
| Risk | T&M | Fixed Price |
|---|---|---|
| Scope uncertainty | easier to absorb through reprioritization | risky without a change request |
| Provider's effort overrun | more often billable when the work is authorized | more often reduces the provider's margin |
| Customer budget predictability | lower | higher for frozen scope |
| Need for clear acceptance criteria | important | critical |
| Legal / liability risk | depends on the contract | depends on the contract |
In the logic of the lecture, Fixed Price becomes realistic when:
If these conditions do not hold, T&M or a phased engagement model is often more transparent.
A proper estimate does not begin with “I think it will take three months”. It begins with decomposition.
scope / assumptions
↓
workstreams / WBS
↓
roles + effort
↓
materials + external services
↓
dependencies / lead times
↓
risk / contingency
↓
commercial estimate
Every significant number needs an explanation: where it came from and which factors would change it.
client need
↓
reference basis
├─ process/framework
├─ standards / legislation
└─ guidelines / project rules
↓
architecture + implementation + V&V
↓
resources + timeline + risks
↓
T&M / FP / hybrid commercial model
↓
Proposal that can be negotiated
An approved proposal leads to another major transition: a contact must become a contract, and the contract must become a delivery team and requirements. The next session starts the journey from a signed agreement through staffing, product discovery, business analysis, user stories and system requirements to the first architecture visualization.