Skip to content
PorkyPhoenix091Public

About

Resources

Stars

0 stars

Watchers

0 watching

Forks

Latest commit

 

History

107 Commits

Folders and files

Repository files navigation

Sorteat

Sorteat — BitBuilders

Human-Computer Interaction team project · Politecnico di Milano · A.Y. 2025/26 Course: Fondamenti di Human Computer Interaction · Final grade: 30/30 cum laude (30/30 e lode)

Handling all the thinking, so you can focus on the eating.

Sorteat is a mobile app concept for shared households: flatmates, couples and families who share a fridge. It keeps track of what is in the house, what is about to expire, who owns what, and what can be cooked with it. The goal is less food waste and less of the mental effort behind it.

Project website · Live high-fidelity prototype · Interface language: Italian


Team

BitBuilders is a six-person academic team:

  • Massimiliano Ferretti
  • Anthony Gerardo Giuliano
  • Sofia Grumelli
  • Jacopo Ascanio Roncato
  • Marco Angelo Ruggieri
  • Mariia Sarbash

The problem

Needfinding showed that household food waste is mainly a problem of disorganisation and fragmented communication, not a lack of interest. Without a shared system, people rely on memory and occasional conversations. This leads to duplicate purchases, forgotten food and waste. Waste also carries an emotional cost, such as guilt and frustration, alongside the financial one. Needs differ between households, but the end goals are the same: reduce waste and save money.


Design process

The project followed six course deliverables. Every document is in this repository and linked from the project website.

# Phase What we did Deliverables
1 Needfinding Semi-structured interviews following a five-part guide, conducted in sub-teams of 2–3 people. 13 participants were contacted across different living situations (alone, with family, shared flats). Interviews were transcribed and analysed thematically. Needfinding (PDF)
2 Refinement & focus Two in-person focus groups (about one hour each) with 6 previously interviewed participants to validate and rank the themes, plus a new round of semi-structured interviews to widen the sample. Refinement & focus (PDF) · FigJam board
3 Tasks, storyboards, first prototypes User tasks and storyboards, then two alternative low-fidelity concepts: a smartphone app and a smartwatch companion. Tasks & storyboards (PDF) · Wireframe #1 (app) · Wireframe #2 (smartwatch)
4 Mid-fidelity prototype Interactive Figma prototype, the design rationale, and a briefing document for external evaluators. Mid-fidelity prototype · Design choices · Document for external evaluators
5 Heuristic evaluation (conducted) BitBuilders carried out a Nielsen heuristic evaluation of another team's project: StrandU, a mid-fidelity Figma prototype by the team Deadline Dodgers. Heuristic evaluation of StrandU (PDF)
6 Redesign, usability testing, high-fidelity prototype Sorteat was redesigned in response to the heuristic evaluation received from the peer team EcoDevs and to feedback from the instructors. Then a high-fidelity prototype was built and tested with users. Prototype correction report (PDF) · Usability testing protocol & script (PDF) · Usability testing report (PDF)

Heuristic evaluation: received and conducted

There were two separate evaluations:

  • Received. Sorteat was evaluated by the peer team EcoDevs. The correction report lists each problem they found, with the Nielsen heuristic it violates and its severity, and explains how the redesign fixed it. One example: the binary "shared / private" toggle became ownership by specific flatmates.
  • Conducted. BitBuilders evaluated StrandU by Deadline Dodgers. That document is about the other team's app, not about Sorteat.

Usability testing

  • 11 participants, all living in shared settings (flatmates or families)
  • Moderated, in-person sessions using the think-aloud protocol, about 45 minutes each
  • 11 sequential tasks covering the whole product flow, with participants playing the user "Mari"
  • Measures: SEQ for each task and SUS for overall usability
  • Average SUS: 89/100 (lowest score 75, highest 100)

Full results, observations and recommendations are in the usability testing report.


The high-fidelity prototype

Open the live prototype → · Source: mariisarbash/sorteat_high_fidelity

A React web app designed for a mobile screen, with four main areas:

  • Home: what's expiring soon, the shopping-list count, product search and an anti-waste streak.
  • Inventario: stock split into fridge, pantry and freezer. Each product has a quantity, an expiry state and one or more owners among the flatmates. A product can be consumed, edited, deleted, or requested from its owner ("Chiedi in prestito", ask to borrow).
  • Spesa: a shared shopping list where each item is assigned to one or more flatmates. A checkout flow reviews the items and moves them into the inventory.
  • Cucina: a meal planner with lunch and dinner slots. Each recipe's ingredients are checked against the inventory as in stock, to buy or ask a flatmate.

Prototype scope. This prototype was built for usability testing and is not a production app. It runs on in-memory mock data that resets when the page reloads, as a single simulated user ("Mari"). Receipt scanning and voice input are simulated.


Repository structure

consegna-1/ … consegna-5/   course deliverables (PDFs and link files)
docs/                       project website (served by GitHub Pages)
├── index.html, style.css
├── images/                 logo and team photos
└── pdf/                    deliverables linked from the website

Signed participant consent forms were collected for the research sessions. They contain personal data, so they are kept privately by the team and have been removed from the current contents of this repository.


Sorteat was designed and evaluated by the six members of BitBuilders as a course project at Politecnico di Milano.

About

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors