SUGGESTORR
rideau rockcliffe community centre suggestion system
SuggestoRR was designed for the Rideau Rockcliffe Community Resource Center (RRCRC). It is an evolution of the fossilized suggestion box, making modern feedback technology accessible for the elderly and youth. The design draws attention, while the technology inside automatically digitizes handwritten suggestions using AI Optical Character Recognition (OCR). It also features a QR code, allowing for digital input. The service end of the system enables communication between new community members and upper management. SuggestoRR bolsters community agency, and in turn, increases community interaction.
Term Project
2026
12 weeks
Carleton University
Skills:
Python
Electronics
Generative Field Research
Wood/metalworking
3D printing
Use Case Marketing Video: Turn on your audio!
01 PROCESS
INITIAL BRIEF
"Design an artifact that helps the community centre gather stories to interpret needs, desires, and opportunities for better programming. "
PHASE 1: GROUP RESEARCH
Expertise
Organizer
Maker
Strategist
Social Researcher
Seller
The research team for this project was built by identifying five individuals with predilections for the listed skills. This created a natural order and an initial methodology for our collaboration.
FIELD RESEARCH - RRCRC
Purpose
"Analyze the RRCRC. Seek to understand it, and when looking for problems view them through a systematic lens."
System To Address - Navigation
The centre is disorienting to navigate as a newcomer. The resulting experience can be discomfort, and feeling estranged. Navigation concerns stretches from physical way-finding to understanding volunteer oppourtunities.
SECONDARY RESEARCH
Here we conducted, micro, macro, financial, and operational background research relating to the RRCRC.
Microeconomic Trends
-
RRCRC exhibits steady financial growth over the past decade.
-
Volunteer onboarding only occurs once per month, during a short time window.
Demographics
-
Largest volunteer age group is 50+.
-
Second largest group is recent immigrants looking to build a network.
-
Third is college students looking to complete mandatory volunteer hours.
-
Large french presence.
-
Multi-cultural.
RRCRC Visitation Data (RRCRC Annual Reports)
-
Despite the number of volunteers nearly doubling from 2022-2024, the volunteer hours contributed only exhibited a marginal increase in community interactions with the volunteers.
-
Repeat interactions from community volunteers have decreased since 2021
-
Though this could be viewed as positive, representing decreased dependency on the community centre survival services...
-
It also represents decreasing community connections.
-
-
How can we reverse these trends, increasing community and volunteer interaction with the centre?
Macroeconomic Trends (The Liquid Society, 2026)
-
Seamless digital interactions integrated in almost every human interaction.
-
People are moving from consumers to creators in search of meaning and connection.
-
Automation is taking over repetitive manual tasks, shifting human labour demands.
RRCRC Current Programming Structure
Poverty Reduction Programs
-
Emergency food bank
-
Mobile market (food truck bank)
-
Social harvest/Chez 13 (farming)
-
Gifts in kind (care packages)
-
Incubator13 (Social Innovation Lab)
Community Support Programs
-
Crisis Support (counselling)
-
Community Development
-
Events
-
Post Incident Support (PINS)
-
-
Adult, senior, and youth programs
-
Tax clinic
Administration and Management Programs
-
Finance
-
Volunteer management
-
Communications and HR
-
Resource Development
Secondary Research Takeaways
To push the RRCRC into the future and allocate resources more efficiently, they would benefit from automation. The solution should accommodate older demographics and people whose native language is not English. Moreover, it should induce repeat interactions for community members and volunteers.
RRCRC INTERVIEWS - KEY INSIGHTS
Volunteer - Ounissa
Ounissa is fluent in French but struggles with English. She joined the RRCRC to network as a recent immigrant. She finds Ottawa winters can be isolating, and struggles to find consistent activities for her kids and herself.
Financial Manager - Sebastien
Volunteers are stretched thin. Background checks and interviews are time-intensive, so we currently conduct volunteer onboarding just one day a month. This results in a vicious cycle.
Director - Luc
Luc affirmed the demographic research and later weighed in on my design concept’s feasibility.
CO-DESIGN 01 - DEFINING VALUES
Through structured activities with our community volunteer, we identified the most important values which should guide the design.
01 Isolation Reduction
Isolation is often self-perpetuated. We value outlets which pull those in need towards belonging in the community.
02 Long-Term Sustainability
We value creation that aligns with long-term desires and values. Systems should be reasonably maintainable due to collective desire.
03 Feedback Loops
No voice should be left unheard. We value connected communication, which creates goals and drives action. Complex systems are required to prevent broken links and build participant trust.
PHASE 2: INDIVIDUAL DESIGN PROCESS
DIVERGENT IDEATION
"Generate 10 concepts that you believe could improve the state of your specific system concern. Be sure to explore service and product-based solutions."
CLIENT PITCH
"Take your top 3 ideas and prepare a pitch. Deliver it to the RRCRC director, a social research master's student, and your ID professor."
Presentation Feedback
03 Exterior Rebranding: "This is a great idea! We definitely need some graphic design done for new window posters."
02 Community Shuttle: "Accessibility is a wicked problem, if you dive deeper into this issue you may end up in-over your head."
01 SuggestoRR: "This is an exciting concept, but it may be too complex given the timeline."
Progression Decision
I was passionate about SuggestoRR and believed in its potential. Following a discussion with my professor, the director, and the RRCRC executive director, it was agreed that I should pursue SuggestoRR due to its potential value.
MARKET ANALYSIS - SUGGESTION & FEEDBACK PRODUCTS
While there is extensive competition in the feedback systems market, SuggestoRR's physical interface, technological integration, and closed-loop social system make it the only competitor in its market space offering all benefit categories.
The market research revealed that SuggestoRR provides novel value to community feedback systems that is not currently offered by competitors.
CO-DESIGN 02 - CLOSING THE FEEDBACK LOOP
SuggestoRR makes providing suggestions a rewarding, enjoyable experience. To close its feedback loop and promote user trust, we wireframed all the links of the communication system. Ounissa (community volunteer) added critical input to the suggestion template, including contact info and a French translation.
People visit the centre
The community votes on management's picks
Interaction increases as people want to get involved
They socialize at the "Ideation Station"
They submit their ideas to SuggestoRR
Front desk sends the suggestion list to management
People get excited about their ideas
Management chooses between suggestions.
Suggestion's project status are updated on the RRCRC website
The community sees their suggestions being implemented.
PHASE 3: PROTOTYPING
STAGE 1: SOFTWARE DEVELOPMENT
Coding with AI OCR, I successfully wrote a script that could extract an image of a filled-out suggestion card from a specified folder, pull the data, and populate the Excel accordingly.
01 Suggestion Card Format
02 Excel Formatting for Data Input
03 VS Coding
STAGE 2: WORKS-LIKE PROTOTYPING
This step was centralized on the mechanics. What insert slot position would allow the suggestion pages to fall into the box in the correct orientation every time?
Findings
The top position generated enough momentum for a page to enter the box cleanly. However, occasionally the page flipped over on its way down. Thus, a high 45-degree entry ensured that the paper settled correctly every time.
To prevent people from inserting their suggestions upside down, I designed a keying affordance. The insertion slot only fits the short end of the page, and the chamfered corner of the page and the slot align. These measures provided only one easy way to insert the page.
STAGE 3: FORM DEVELOPMENT
SuggestoRR had to be intriguing yet approachable, and fitting within the community centre context.
STAGE 4: HARDWARE, FIRMWARE, AND SOFTWARE SYNCING
Due to the porous properties of paper, the break-beam sensor would not register being broken if the IR signal was too strong. Blocking the receiver proved to be more effective than blocking the IR sender. These factors informed the positioning of the break-beam modules in the final design.
STAGE 5: FINAL PROTOTYPE
Refining the technical drawings took over a dozen iterations. The interplay between physically building the model and updating the drawings accelerated product development rapidly.
Building Stages
1. Planing, mitre-sawing
2. Milling, drill-pressing
3. Routering, sanding, inserts
4. Wood glueing
5. Varnishing
6. Metal shearing
7. Steel punching
8. Break-bending
9. Custom mold bending
10. Assembly
01
02
03
04
05
06
07
08
09
10
02 FINAL DESIGN
SuggestoRR was a success. The hardware and firmware worked flawlessly, Even with precise revisions, it is 65% faster than manually digitizing suggestions, while providing the added benefits of time-stamping and recording visuals.
65%
Faster Documentation
95%
Sentiment Accuracy
>4x
Interaction Frequency
I conducted a controlled study at the RRCRC measuring the difference in interaction frequency between SuggestoRR and a traditional suggestion box. The results concluded that the SuggestoRR is intriguing, collecting 4 times more suggestions than the traditional suggestion box in a fraction of the time.
MATERIALS
Red Oak Hardwood
Walnut Gel Stain
Polyurethane
SB Steel
Black Fasteners
Clear Acylic
The plug tucks neatly away during transportation, and the handles provide a reliable affordance. The side compartment stores pens and blank suggestions.
Rubber feet provide stability, non-slip, scratch prevention, and cleanliness. The front sticker provides an affordance security beyond the design, dictating how to use SuggestoRR. On top, there is a QR code linked to a digital form with identical questions to the physical cards. This offers a second input outlet for those who prefer to save paper.
SuggestoRR components can be easily replaced, and the raw materials are separable for recycling. The side panel is removable for tech servicing.
At the end of each day, the desk clerk shuts down SuggestoRR. They would then review the suggestions and affirm that SuggestoRR made no major errors in the sentiment analysis. As it collects data, its algorithm will improve.
SOUND ON
HOW IT WORKS
SuggestoRR creates an outlet for community members to engage with the centre, each other, and RRCRC management. It uses existing RRCRC resources: The product is overseen by the directory desk staff. There is already a program for community-led initiatives, so that department would oversee executive decisions for the SuggestoRR program. The RRCRC already has a website, so the web designer would update the SuggestoRR section. While it incorporates automation, it is the human touch that makes the product desirable through shared involvement.
What the Camera Captures:
People visit the centre
Interaction increases as people want to get involved
They socialize at the "Ideation Station"
They submit their ideas to SuggestoRR
Front desk sends the suggestion list to management
People get excited about their ideas
Management chooses between suggestions.
Suggestion's project status are updated on the RRCRC website
The community sees their suggestions being implemented.
The community votes on management's picks
Working Prototype Test: Leave on your audio!
CURCUIT SCHEMATIC
While it is somewhat crude to use a breadboard in a prototype, rather than soldering all connections, it allowed me to complete the circuit without purchasing extra components or wire splitters. The real product would have a custom PCB with only the necessary components.
Break-Beam IR Reciever
Break-Beam IR Emitter
Breadboard
RPi USB-C Power Supply
5V RGB LED Strip
Camera
Module 3
RPi Model 4B
DATA FLOW
Pictured below are the automated stages for data transfer, from paper to Excel. The Syncthing folder is shared between the RPi and the laptop. My laptop runs a watchdog, which notices when new suggestions are uploaded to the folder. It notifies the script, which extracts the new suggestion and runs it through AI-OCR, recognizing the fields and text. Lastly, it stores the data in the next available row in the master Excel.
Syncthing Shared Folder
Suggestion Card
Camera
Excel Spreadsheet
RPi Memory
VS Code Script
THE LAST 10%.
