Chapter 49 — Engineering thesis
Lesson map
Lesson 49 published. Map, example, EX097.
flowchart TD
N0["Learning outcomes and connection"]
N1["49.1 The engineering thesis"]
N2["49.2 Structure"]
N3["49.3 Results and limitations"]
N4["49.4 Writing well"]
N5["49.6 Practical activity"]
N6["Chapter synthesis"]
N0 --> N1 --> N2 --> N3 --> N4 --> N5 --> N6
N6 --> EX["Worked example"]
EX --> Q["Exercise EX097"]
Logical drawing
flowchart LR
N0["Learning outcomes and connection"]
N1["49.1 The engineering thesis"]
N2["49.2 Structure"]
N3["49.3 Results and limitations"]
N4["49.4 Writing well"]
N5["49.6 Practical activity"]
N6["Chapter synthesis"]
N0 --> N1 --> N2 --> N3 --> N4 --> N5 --> N6 --> Q["EX097"]
[Learning outcomes and connection]
|
v
[49.1 The engineering thesis]
|
v
[49.2 Structure]
|
v
[49.3 Results and limitations]
|
v
[49.4 Writing well]
|
v
[49.6 Practical activity]
|
v
[Chapter synthesis]
|
v
[Worked example] --> [Exercise EX097]
Learning outcomes and connection
After this chapter you can (1) structure an engineering thesis around a design problem, solution and evidence; (2) write results and limitations that stay within the evidence; (3) present quantitative results with uncertainty and provenance; and (4) prepare a thesis that an examiner can verify. It turns the dossier into a document and prepares the defense.
49.1 The engineering thesis
An engineering thesis argues that a designed system meets a need under constraints, supported by verification and validation evidence. It is not a research dissertation (the master's book covers that) and need not claim an original contribution to knowledge (the doctoral book). Its originality, if any, lies in the design and the quality of its evidence.
For TMRP, the thesis argues that the restart-safe gateway meets R1, R2 and R7 on the bench and reports R6 as not met.
49.2 Structure
- Introduction: need, stakeholders, consequence, scope.
- Background: sector context, relevant standards and prior designs, with sources.
- Requirements: derived requirements with sources and verification methods.
- Design: alternatives, selection, architecture, controls, with decision records.
- Analysis: threat and failure analysis, performance, reliability.
- Implementation: what was built, how, with which tools.
- Verification and validation: methods, results, failures and fixes.
- Discussion: requirement status, residual risks, limitations, what you would do differently.
- Conclusion: what was achieved and what remains. Appendices: traceability matrix, dossier index, lab code.
The TMRP thesis follows this structure.
49.3 Results and limitations
Report requirement status honestly: met, met after fix, not met, not verified. For each limitation, say which claims it affects and in which direction: "the bench uses perfect clocks; real clock drift would increase stale-message rejections, so the false-stale rate reported is a lower bound". A limitation without a direction is decoration.
The TMRP discussion paragraph in the worked example below gives such a limitation: the bench used a synthetic network, so the sixty-second stale-station threshold must be re-tuned on site, where LTE congestion may raise false alarms. That sentence names the affected claim and the likely direction.
49.4 Writing well
Lead each chapter with its main point. Use figures with labelled axes and units; tables for exact values. Keep numbers traceable to the dossier. Write in plain technical English, define terms at first use, and use a glossary.
For TMRP, the requirement-status table appears early in the discussion, and every number in it traces to an entry in the dossier index with its hash. Each chapter opens with its main point, and terms such as stale-station alarm are defined at first use and in the glossary, so an examiner can read any chapter without hunting for definitions.
49.5 Worked example: discussion paragraph (synthetic)
The gateway met R1 and R2 on the bench, including restarts of either end, after the receiver was changed to persist its last accepted counter (Section 7.3). R7 was met for priority alarms and, after review finding F2, for loss of data. R6 was not met: the recovery trial restored remote operation in 12 h against an 8 h requirement, because forensic imaging and rebuild lie in series on the critical path. Hospital supply was maintained in local mode throughout the trial, so the consequence of the shortfall is reduced visibility and intervention capability, not loss of supply. Two design options that would meet 8 h are costed in ADR-015. All results come from a synthetic bench; field performance, particularly LTE latency under congestion, may differ, and the 60 s stale-station threshold must be re-tuned on site.
49.6 Practical activity
Write the discussion chapter of your thesis, with a requirement-status table and directional limitations.
Use the TMRP discussion paragraph as your model. Begin with a table that gives each requirement's status as met, met after fix, not met or not verified. Then write one paragraph per significant result, including any requirement you did not meet, and end with your limitations, each naming the claim it affects and the direction of likely bias. Check every number against your dossier index before submitting.
49.7 Common errors and limits
- Theses organised as diaries.
- Requirement status rounded up.
- Limitations without direction or affected claims.
- Numbers that do not match the dossier.
The TMRP thesis avoids each: it is organised by argument, reports R6 as not met, gives directional limitations, and draws its numbers from the dossier index. Its requirement-status table is checked against the dossier before submission, so nothing is rounded up. Its limitations name the synthetic bench and the single recovery trial, and say in which direction each might bias the results.
Exercises — chapter 49
EX097 — Directional limitations. Objective: write useful limitations. Inputs: three limitations: synthetic noise model; bench without real LTE; one recovery trial. Steps: for each, state the affected claim and direction of likely bias. Expected evidence: three sentences. Rubric: 3 = each has claim and direction (e.g., one trial gives no variability estimate — 12 h could be optimistic or pessimistic; real LTE latency would increase stale rejections and alarm latency); 2 = one missing direction; 1 = list only; 0 = none.
EX098 — Abstract. Objective: summarise accurately. Inputs: §49.5. Steps: write a 150-word structured abstract (need, design, methods, results, limitations). Expected evidence: abstract. Rubric: 3 = accurate including R6 not met, synthetic-bench limit stated; 2 = one overstatement; 1 = vague; 0 = none.
Chapter synthesis
An engineering thesis argues from need to design to evidence. Structure it for an examiner, report requirement status as it is, give every limitation an affected claim and direction, and keep every number traceable to the dossier.
Continue studying — chapter 49 of 50
← Chapter 48 · Course chapter index · Chapter 50 →
Continue to the next authored lesson.