Team and responsibilities
Four roles, who does what, effort and working practices.
The project needs four people with complementary skills. This chapter defines the four roles, what each person is responsible for, how effort is distributed and how the team works together. Names are entered when the positions are filled; throughout the report the roles are referred to by their codes.
10.1Roles
Table 10.1.The four roles.
| Code | Role and profile | Responsibilities | Ownership |
|---|---|---|---|
| PI | Principal investigator (physicist) Electromagnetic modelling, measurement design and data analysis; scientific Python; laboratory practice. | Scientific direction and project management (WP0). Forward models and calibration against exact solutions (WP2). Tensor inversion and radar analysis (WP4, WP5). Fusion method, field-trial design and publication (WP6). Chairs reviews and decision gates. | Leads WP0, WP2, WP6 13 tasks Owns D0.1, D2.1, D3.2, D4.2, D5.2, D6.2 |
| HW | Hardware engineer (analog and RF) Analog circuit design, PCB layout, coil winding, RF measurement with a vector network analyser, mechanical fabrication. | Coils, transmit drivers, preamplifiers and nulling (WP1, WP2). Sensor pole mechanics (WP3). Three-axis coil head (WP4). Antennas, amplifiers and radar front end (WP5). Safety officer; places objects for blind tests. | Leads WP1, WP5 11 tasks Owns D1.2, D2.2, D5.1 |
| SW | Embedded and software engineer Embedded C/C++, real-time acquisition, Linux, Kalman filtering, software engineering practice. | Codec bring-up and host streaming (WP1). Acquisition software. GNSS, IMU and timing integration and the fusion filter (WP3). Radar sweep software (WP5). Live map, data pipeline, repository and public release (WP6). | Leads WP3 10 tasks Owns D1.1, D3.1, D6.3 |
| DS | Data scientist and field engineer Machine learning, statistics, simulation, experimental design; field work. | Test objects and acquisition support (WP1). Ground survey (WP2). Camera odometry (WP3). Simulator, dataset collection, classifiers and blind evaluation (WP4). Sensor fusion and dataset release (WP6). | Leads WP4 7 tasks Owns D4.1, D4.3, D6.1 |
10.2Responsibilities by work package
Each work package has one accountable lead and at least one second person who reviews and can take over. The matrix uses the usual RACI convention: R responsible (does the work), A accountable (owns the result and signs off), C consulted, I informed.
Table 10.2.Responsibility matrix by work package.
| WP | Title | PI | HW | SW | DS |
|---|---|---|---|---|---|
| WP0 | Project management | A/R | C | R | C |
| WP1 | Bring-up and receive chain | C | A/R | R | R |
| WP2 | Static induction rig and calibration | A/R | R | C | R |
| WP3 | Positioning and timing platform | R | R | A/R | R |
| WP4 | Full tensor, dataset and classification | R | R | C | A/R |
| WP5 | Radar subsystem | R | A/R | R | C |
| WP6 | Fusion, field trials and release | A/R | R | R | R |
10.3Effort
Each person works full time for twelve months, giving 48 person-months in total. The table shows the planned distribution. It is used to cost the work packages in Chapter 11.
Table 10.3.Effort in person-months.
| Role | WP0 | WP1 | WP2 | WP3 | WP4 | WP5 | WP6 | Total |
|---|---|---|---|---|---|---|---|---|
| PI | 1.5 | 0.5 | 3.0 | 1.0 | 2.0 | 1.0 | 3.0 | 12.0 |
| HW | 0.3 | 3.0 | 2.5 | 1.2 | 1.5 | 3.0 | 0.5 | 12.0 |
| SW | 0.3 | 2.5 | 0.5 | 4.5 | 0.5 | 1.7 | 2.0 | 12.0 |
| DS | 0.2 | 0.8 | 1.5 | 1.5 | 5.0 | 0.5 | 2.5 | 12.0 |
| Total | 2.3 | 6.8 | 7.5 | 8.2 | 9.0 | 6.2 | 8.0 | 48.0 |
10.4Focus by quarter
Table 10.4.Main activities of each role per quarter.
| Role | Q1 (weeks 1–13) | Q2 (14–26) | Q3 (27–39) | Q4 (40–52) |
|---|---|---|---|---|
| PI | Forward model, calibration plan, procurement | Calibration against theory, M1 review | Tensor inversion, radar analysis | Fusion, field trials, paper |
| HW | Coils, preamplifiers, TX driver | Nulling, pole mechanics, tensor head | Radar antennas and amplifiers | Field hardware, documentation |
| SW | Codec bring-up, acquisition software | RTK, timing, sensor fusion | Radar software, data pipeline | Live map, release |
| DS | Acquisition software, test objects | Ground survey, camera odometry | Simulator, dataset, classifier | Blind test, fusion, dataset release |
10.5Working practices
- Two people per result. Every measurement or analysis that feeds a milestone is checked by a second team member before it is reported.
- Lab log. Each measurement session is recorded with date, set-up, settings, data file names and observations, including failed attempts.
- Code review. All code is merged through reviewed pull requests in the project repository; analysis scripts are versioned with the data they produced.
- Decision log. Design choices and gate decisions are recorded with the alternatives considered and the reason for the choice.
- Blind testing. The person who buries test objects for an evaluation (normally HW) does not take part in the analysis of that test.
- Safety officer. HW is responsible for the electrical and field safety rules in Chapter 12 and briefs each new participant.
10.6Substitution
If one person becomes unavailable, the second owner of each of that person's tasks (the RACI "R" entries of another role) continues the work. For PI, the DS role takes over analysis and SW takes over coordination. For HW, PI covers measurements and an external contractor may be engaged for PCB work (contingency budget). The written lab log and decision log make such transfers possible without loss of information.