Projects on Josh M

Robotic Fixture Design

Mon, 01 Jan 0001 00:00:00 +0000

See it in action

Your browser does not support the video tag.

Gallery

1 Fixture Plans Overview.png
2 Fixture Plan Detail.png
3 Plc Programming.png
4 Pneumatic Diagram.png

Situation

Problem: Manual clamping and welding processes were labor-intensive and physically demanding.
Goal: Automate the part-holding process to improve throughput and workplace safety.
Context: Operators were reporting frequent back injuries due to repetitive, awkward posturing during manual welding.

Task

Responsibility: Lead designer for the mechanical fixture and the control logic.
Constraints: Must integrate seamlessly with existing robot arms and fit within the current cell footprint.
Tech Stack: CAD software for modeling, PLC for sequencing, and pneumatic actuators.

Action

Design: Developed a custom pneumatically actuated clamping fixture using CAD to ensure precision alignment.
Programming: Wrote PLC logic to synchronize sensors, actuators, and robot arm movements for a seamless handshaking sequence.
Prototyping: Collaborated closely with the tool and die shop to fabricate the hardware.
Iterating: Conducted stress testing and refined the clamping pressure and sensor positioning to eliminate part slippage.

Result

Efficiency: Halved labor costs by transitioning from manual to automated operation.
Safety: Significantly improved ergonomics, resulting in a measurable reduction in reported back injuries.

School Behavior App

Mon, 01 Jan 0001 00:00:00 +0000

Situation

A school district needed a centralized way to track behavioral data for hundreds of students across multiple classrooms. Existing paper-based or disjointed systems made it difficult to identify long-term trends, manage reward “point” economies, or keep parents updated in real-time without significant manual effort from teachers.

Task

I was tasked with building a low-friction, high-impact web application that would:

Empower Teachers: Quick entry for positive/negative behavioral incidents.
Engage Students: A portal to view “merit points” and redeem them for school rewards.
Inform Parents: Automated, easy-to-read progress reports.
Equip Admins: High-level analytics to monitor trends per teacher or student over time.

Action

I leveraged Google Apps Script as a serverless backend to minimize deployment friction and technical overhead for the district:

Georgian Morphological Segmentation & Visualization

Mon, 01 Jan 0001 00:00:00 +0000

Situation

Georgian verbs are famously complex, featuring a multi-slot morphological system where a single word form can encode subject, object, and version markers. Manually analyzing these forms is time-consuming, and standard text-based dictionaries fail to illustrate how individual morphemes shift across different grammatical categories.

Task

I was responsible for building a technical bridge between raw linguistic data and human-readable analysis:

ML Segmentation: Develop a model to automatically split verb forms into constituent parts (e.g., ga-v-aket-eb-di).
Interactive Visualization: Build a web-based dashboard to align and compare these segments visually for researchers and students.

Action

I approached this as a two-stage engineering problem:

Automated Invoicing Pipeline

Mon, 01 Jan 0001 00:00:00 +0000

Gallery

1 the Complex Vendor Form.png
2 the Returned Vendor Data.png
3 Results.png

Situation

The client’s quoting process was stalled by a massive manual bottleneck involving:

Paintline Optimization

Mon, 01 Jan 0001 00:00:00 +0000

Situation

Problem: Inefficient powder utilization and frequent, unoptimized color changes were driving up material costs and reducing throughput.
Goal: Increase powder transfer efficiency and streamline the scheduling of parts on the moving track.
Context: The system utilized spray guns to coat charged parts on metal hangers across a continuous conveyor line.

Task

Responsibility: Lead data collection, nozzle calibration, and the development of a digital scheduling solution.
Constraints: Must work within part batch deadlines and provide an accessible interface for floor managers.
Tech Stack: VBA (Excel-based tool), ultrasonic thickness gauges, and flow control sensors.

Action

Data Analysis: Audited paint thickness across various part geometries and colors to identify areas of over-application.
Process Tuning: Optimized spray nozzle outputs to ensure uniform coverage, reducing wasted powder and “orange peel” defects.
Software Development: Built a custom VBA tool that allowed managers to input part batches and constraints.
Optimization: Programmed the tool to minimize high-cost color changes and part-size transitions while meeting strict delivery timeframes.

Result

Financial Impact: Netting $20,000+ in annual savings through reduced powder waste and improved labor efficiency.
Consistency: Achieved high-uniformity coating standards, reducing rework by stabilizing paint thickness.

Drafting Automations

Mon, 01 Jan 0001 00:00:00 +0000

Gallery

1 Wrote Code to Automatically Mark Clear Points.png
2 Wrote Code to Automatically Draw Where Underground Utilities Crossed Rail Alignment.png

Situation

Problem: Manual drafting workflows for large-scale civil projects were prone to human error and consumed excessive billable hours.
Goal: Create a robust set of “one-click” tools to handle complex geometric exports and bulk text edits within the AutoCAD environment.
Context: Specialized tasks like railway turnout extraction and coordinate data exporting required custom logic not found in the standard Civil3D toolkit.

Task

Responsibility: Design, code, and document a collection of productivity scripts to be used by the wider drafting team.
Constraints: Tools had to be compatible with legacy AutoCAD versions and support external data formats (CSV/TXT) for seamless integration with field surveys.
Tech Stack: AutoLISP for core CAD manipulation and VBA/Instructions for data management.

Action

Data Automation: Developed scripts to automatically draw and export point data, eliminating manual entry and ensuring coordinate accuracy between the model and field files.
Geometric Specialized Tools: Built a custom Turnout Data Export utility to handle specialized rail geometry calculations and reporting.
Bulk Processing: Created an Incremental Text Replacement tool to automate sequential labeling and systematic text updates across entire drawing sets.
Documentation: Authored comprehensive Work Instructions for each utility to ensure consistent application and reduce the learning curve for new users.

Result

Speed: Reduced data extraction and export times by approximately 80% compared to manual methods.
Accuracy: Eliminated transcription errors by automating the flow of data from the drawing space to external utility reports.
Scalability: The tools allowed the team to handle significantly larger datasets without increasing the drafting headcount.

Biomass Gasification Research

Mon, 01 Jan 0001 00:00:00 +0000

Situation

The Problem: Biomass and municipal solid waste (MSW) are underutilized energy sources due to high tar production and low conversion efficiency.
The Goal: Optimize a 75 KWth bubbling fluidized bed (BFB) gasifier to handle diverse feedstocks (industrial/agricultural waste).
Environment: Research conducted in the on-site bay using a system funded by a $60,000 NSF EPSCoR grant.

Task

Design: Integrate a biogas burner into the existing gasification infrastructure.
Analyze: Evaluate how feedstock variation and reactor hydrodynamics impact cold-gas efficiency.
Maintain: Ensure project continuity through rigorous data documentation for faculty and future researchers.

Action

Hardware Engineering: * Designed and installed improvements to a biogas burner as part of a collaborative engineering team.
Experimental Testing: * Conducted parametric experiments focusing on temperature, bed composition, and fluidization hydrodynamics while analyzing gas chromatography.
Data-Driven Optimization: * Identified correlations between feed system configurations and tar production levels.
Knowledge Management: * Drafted technical reports and findings for the Engineering Department Chair.

Result

Efficiency Gains: Identified specific operating parameters that maximized cold-gas efficiency while minimizing tar.
System Capability: Successfully upgraded the research platform with a functional burner for synthesis gas combustion.
Academic Impact: Created a foundational dataset and documentation library used by subsequent student research teams.