Lior Meir

The Dash is an approximation of a race-engineering workspace. It keeps the live session state visible and adds analytical layers where they help frame what may happen next. Publicly available data defines the boundary; the project is not intended to reproduce a team pit wall or infer information that is not available.

The design principle is to surface information and relationships that support engineering interpretation rather than automate it. Development therefore centres on defining a useful calculation, testing it against live race-session behaviour, and revising the logic when its assumptions fail.

Project contribution centres on engineering and analytical definition: the questions the Dash should answer, calculation and metric logic, assumptions, information hierarchy, live testing, interpretation of failures and development priorities. Python and Streamlit implementation is AI-assisted; the application is not presented as evidence of independently authored software.

External inputs include accessible Formula 1 data streams and researched values such as circuit-specific pit-loss estimates. These inputs are separated from the analytical logic developed around them.

A plausible lap time is not enough to establish that the model is behaving correctly. Errors can cancel: a model can be too fast in one phase and too slow in another while still producing a convincing total time. The more useful question is therefore where the predicted speed trace agrees with the reference, where it diverges, and what part of the model could physically create that difference.

The project therefore uses differences in trace shape and regional speed trends as diagnostic evidence rather than relying on a headline lap-time number. A speed difference may come from the vehicle model, tyre assumptions, represented track curvature or driven line, imperfect dataset alignment, or several of these effects together.

Authorship is mixed. Vehicle-dynamics reasoning and modelling direction are combined with academic and self-directed learning, AI discussion, methodological guidance and AI-assisted implementation. Development starts from a physical interpretation, then uses discrepancies to identify what needs to be learned or revised next.

The Lap-Time Simulator is presented as evidence of an active vehicle-dynamics learning and modelling process, not as a claim of an independently authored or professionally validated simulation package.

The value of the exercise is the requirement chain rather than the absolute aerodynamic numbers. It connects circuit geometry and observed vehicle speed to a vehicle-level load requirement before later aerodynamic development, rather than beginning from wing coefficients in isolation.

The calculation also makes its assumptions visible. Large or unexpected requirements can indicate either a demanding circuit condition or a limitation in the simplified vehicle, tyre, corner-radius or load-distribution assumptions.

The contribution presented here is the first-sheet requirement study: circuit assumptions, force and downforce calculations, and rear-wing requirement derivation.

Candidate-wing assessment and the subsequent detailed DRS and airfoil development were carried out within the wider group project and are not presented here as personal work.

The branched internal passage created the key tooling problem. A core with the correct negative volume is not useful if it cannot be released from the tooling used to make it. The solution was therefore a deliberately sequenced two-piece core system rather than a monolithic branched shape.

The branch piece is withdrawn first along the branch direction. Removing it clears the geometric interlock, allowing the main core to withdraw along the primary passage axis. Core prints then locate the assembled core relative to the external mould cavity.

This was a group academic project. The contribution presented here covers component interpretation and reconstruction, CAD development, the as-cast transformation, pattern and mould geometry, and the core and core-tooling solution. Pouring-system development, including gating, runners, riser and venting, was primarily completed by other group members and integrated into the wider manufacturing package.

The project is presented as evidence of design-for-manufacture and geometric reasoning, not as production-foundry tooling. Process allowances and parameters were selected from researched casting guidance within the scope of the assignment rather than validated through a production casting programme.

The main challenge was identifying what combination of curves, sections and surfaces could plausibly generate the observed geometry. With only photographs available, the construction method had to be inferred alongside the shape.

This is a CAD and geometric-interpretation study only. The profiles were not derived from aerodynamic requirements and the model is not presented as an accurate reproduction of a specific IndyCar aerodynamic surface.

The main challenge was systems integration: the mechanical structure, actuation, sensing and control had to operate as one physical system. Development began by separating the required task from capabilities that a more general robotic manipulator could provide but the brief did not require.

The operating environment used a defined detection and pickup position with fixed destination buckets. Because arbitrary target coordinates were not required, calibrated joint poses and a repeatable motion sequence matched the task; inverse kinematics would have addressed a more general problem than the brief required.

The project is presented as evidence of multidisciplinary systems integration, not advanced robotics. The fixed-location strategy used calibrated joint positions because the required task did not call for arbitrary target coordinates or a general inverse-kinematic solution.