Improved inspection throughput 10% on a 2M+ parts/year production line 

GPA 97/110, ranked 1st on Italy's national engineering entrance exam 

Battery Lead behind 3 consecutive IHPVA World Records 

Set in Nevada on Sep 12, 2024, the fastest arm-powered vehicle record.

Set at Balocco on Oct 18, 2025, as part of Cerberus's arm-powered record run.

Set at Balocco on May 23, 2026, completing Cerberus's third consecutive world record.

• Designed a 12V LiFePO4 battery pack to power Pegasus's onboard electronics

• Powered gearbox controllers (Omero/Vertigo), position-feedback board (Refuso), telemetry, and vision system

• Benchmarked 26650 and 18650 LFP cell formats against capacity, discharge current, and cost

• Selected a 4S configuration to reach the required 12V nominal output (3.4 x 4)

• Split power into a 2S pack for Telemetry/Whereami and a 4S pack for Omero/Vertigo/Refuso to isolate sensitive signals from motor circuits

• Modeled the pack enclosure in Siemens NX

• Kept telemetry and GNSS signal cabling electrically isolated from high-current gearbox circuits

• Reduced electromagnetic interference risk across the CAN bus network

• Balanced capacity, discharge rate, and cost in the final cell choice

• Designed the front fork assembly connecting to the driving wheel of Pegasus

• CAD-modeled and assembled the full fork sub-system in SolidWorks

• Produced the 2D drawing with GD&T for the fork-to-hub joint for production

• Modeled structural parts (clamps, brackets, steering knuckle, column) in Ergal (7075 Al-alloy) for strength, and the handlebar in carbon fiber for weight savings

• Assembled all parts, checking fit and interference between components

• Generated orthographic views, sections, and GD&T callouts (flatness, parallelism, profile tolerances) per ISO 2768-mK and ISO 8015

• Delivered an assembly-ready fork combining high-strength Ergal and lightweight carbon components

• Produced a fully toleranced, manufacturing-ready drawing ensuring proper fit with adjoining components

• Provided the mechanical basis for further structural validation (FEM/topology optimization)

• Optimized the fork joint (steering-to-fork connector) for Pegasus, designed for SLM 3D printing

• Reduced part mass while maintaining structural safety under load

• Defined design space in SolidWorks and ran topology optimization in Altair Inspire

• Switched material from Al7075-T6 to AlSi10Mg (SLM-compatible)

• Validated with FEM under a 1000N load, simulating full rider + prototype weight

• Cut mass from 216.3g to 59.79g, a 72.36% reduction

• Maintained a safety factor above 1.5 with stresses well below yield strength

• 3D-printed a prototype for fit-checks prior to SLM production

• Ran FEM analysis on the Cerberus chassis as a sandwich carbon-PVC composite structure under pedaling loads

• Assessed composite ply failure risk across the frame's carbon fiber layup

• Applied pedaling load case to simulate real riding conditions in Hypermesh

• Evaluated results using the Tsai-Wu failure criterion for maximum ply failure across the composite layup

• Maximum composite failure index reached 2.126E-02, far below the 1.0 failure threshold, indicating a large safety margin

• Confirmed the chassis layup safely withstands combined bending-axial-torsion pedaling loads without static ply failure

• Validated structural integrity ahead of Cerberus's world-record attempts

• Laminated the wheel hubs in carbon fiber for Cerberus, following composite manufacturing standards

• Ensured consistent fiber orientation and resin distribution for structural strength

• Learned fluorescent magnetic particle inspection (MPI) to detect cracks on forged transmission components

• Performed dimensional inspection of finished shafts against engineering drawings

• Applied fluorescent magnetic solution and inspected parts under UV light, where crack indications glow due to magnetic flux leakage

• Flagged forged shafts showing crack indications before further processing

• Measured critical dimensions with a height gauge and cross-checked against GD&T callouts on the drawing

• Gained hands-on exposure to NDT crack-detection methods used in OEM quality control

• Helped prevent defective forged parts from advancing in production

• Practiced linking dimensional measurements to GD&T requirements on real production drawings

• Simulated open-die upsetting of a cylindrical billet using QForm forging simulation software

• Analyzed stress distribution and barreling behavior during compression

• Modeled the billet and die geometry in QForm and ran the compression simulation

• Tracked mean stress evolution and barreling as the billet deformed between the dies

• Captured the classic barreling pattern expected in frictional upsetting

• Validated peak stress concentration at the billet's equatorial bulge, consistent with forging theory

• Completed simulation with full solver convergence

• Delivered a fully assembled, functional cast aluminum hog trophy meeting all specified GD&T requirements

• Demonstrated integration of four distinct manufacturing processes (casting, CNC milling, CNC turning, laser cutting) into one cohesive assembly

• Validated that base thickness, hole location, and thread engagement provided secure mechanical assembly and consistent part-to-part fit

• Purity control drove a 43.9% capacity gain at −20°C, confirmed in 18650 cells

• Optimal carbon coating gave a 58.2% capacity gain at −10°C and raised the TR trigger temperature by +43°C

• Combined strategy reached ~3x baseline capacity at −20°C

• Concluded LFP/carbon is the strongest near-term path; SSBs remain the long-term solution pending cold-conductivity breakthroughs

• Modeled the finished part in SolidWorks, then applied machining allowances and thermal shrinkage to derive cavity dimensions

• Designed the full cavity cross-section (land, flash, gutter) with fillet/draft callouts, breaking it into discrete volumes (V1-V7) for load calculations

• Organized steel grades by classification and chemistry (C, Si, Mn, P, S, Cr, Ni, Mo) using standard designations

• Cross-referenced each grade's annealed and quenched hardness and tensile strength values into one unit-standardized lookup table, eliminating the need to toggle between multiple disconnected data sources