A broad range of City of Melbourne employees who work in open space planning, parks and city greening, urban design and more participated in a workshop. We asked attendees to share experiences of living, working and playing in the City of Melbourne, either personal experiences or known experiences representing other people or groups that they had learned through project work.

This led to the selection of three large areas, approximately 500 metres by 500 metres, where we pursued further insights via data analysis. These areas were classified as north-east Carlton, central Carlton and lower central CBD.

Data analysis begins with compiling data, and for this project, there were two data sources; City of Melbourne’s data from pedestrian sensor network (you can see the data firsthand on City of Melbourne’s Pedestrian Counting System) and purchased data from telecommunications companies. Both data sources contained only de-identified data. The data is created when WIFI and Bluetooth devices periodically broadcast messages that indicate their presence or participation in a network. This helps us understand people movement in our city spaces.

Altometer have been turning data into insights across three City of Melbourne locations. This process revealed behaviour patterns and further sites of interest.

The red areas indicate high levels of pedestrian movement (see GIFs below), orange represents moderate pedestrian movement and green represents low pedestrian movement. Select the GIFs to see movement data animation.

The project team analysed the data and considered other factors including lockdown periods and physical landmarks in order to select the next three key focus sites for this project.

Three sites were selected from this phase. They demonstrated moderate pedestrian activity (orange colour) and the physical spaces exhibited potential for exploration and testing. We also considered things like existing furniture, pedestrian flow, cars and parking, shade, plants and lighting.

Place audits enable insights into how public spaces work at a local level. By capturing stories about how people are engaging with our spaces, we can be better informed to plan, design and maintain them.

Inhabit Place deployed seven auditors to three sites in Melbourne: Macarthur Square, Drummond Street and Argyle Square to capture observational data.

In one day of auditing, the Drummond Street site was the most popular with 2002 people observed. However, unlike Macarthur Square and Argyle Square, 95 per cent of users were passing through with only five percent spending time in the green median strip.

Key user profiles observed across the sites include:

• Macarthur Square: locals
• Argyle Square: workers on lunch breaks, office meet-up
• Drummond Street: shopper, lounger

When auditors interviewed users, they discovered a variety of insights around how they feel at these sites, as well as what improvements they’d like to see here.

The following criteria were applied to determine the final site for the new furniture:

• Is there enough pedestrian activity to test an intervention?
• Did users demonstrate an intention to congregate in the area?
• Does the space provide good opportunities for people to congregate (i.e. enough space, good traffic conditions etc.)?
• Did visitors express other desires or needs during interviews?
• Could the design developed for this site be used for other sites around Melbourne?

Other areas considered in the furniture designs include:

• Transport behaviours and any potential for positive impact
• Existing furniture and the amenity it provides
• Perceptions of safety
• Macro opportunities (i.e. is it a thoroughfare between significant hubs?)

The Drummond Street median in Carlton was determined as the final site.

Get a snapshot of the movement paths in the images below or download our place audits to dive into the data:

• Macarthur Square
• Argyle Square
• Drummond Street

All stages of the design process were informed by the data collected in the earlier phases.

Concept designs demonstrated various ways data could be interpreted into furniture items. See a snapshot of concepts below. Several collaborative workshops developed the data and concepts into a final design.

The following points show how data led to seven design intentions and the outcomes that you can see in the final design. The final design is approximately seven meters long, two and a half meters wide and two meters high. It is constructed from Australian hardwood timber (spotted gum) with an oil finish with a structure made from powder coated galvanised mild steel. Designers considered tree locations, future planting and sprinklers.

See some of the concept designs in the images below.

Design intention 1: Enabling a bustling ‘social space’

Through conversations we learned that users wanted items that can enable busking, live performance or other activations. The design contains a small 'stage' area as well a variety of different places for people to dwell and invite social activation.

Design intention 2: More places to sit and eat

People expressed a desire for more seats and tables for picnicking and “enjoying the outdoors”, to “enliven the space” and provide options for older groups who found sitting on grass inconvenient. Top observed behaviours were dwelling (52%) and eating or drinking (44%) with people sitting on the grass (sun and shade), in nearby alcoves and event leaning on bike racks.

In the design, there are several different places to sit, eat or drink across four distinct features: stage, high bench, double open seat and seat with armrest.

Design intention 3: Safe and share-able

People expressed a desire for furniture that enabled people to maintain social distance. A long design (seven metres) allows users to maintain social distancing. The spine of the design provides some privacy between the two sides while maintaining visual connections through the perforations.

Design intention 4: Efficient and useful

The design can accommodate the following activities at the same time: a person can be sitting on their own eating a picnic lunch, children could be playing on the adjacent stage, friends could be catching up on the other side of the chair, with a couple of colleagues standing and using the ledge to eat lunch and catch-up. This efficient and useful design responds to user feedback that the space feels underwhelming, underutilised and segregated.

Design intention 5: Supporting diversity

Diverse activities included work breaks (uniformed users), family outings, leisure and exercise, dog-walking, grocery shopping, hospitality patrons and more contributed to a sense of ‘buzzing’.

The design expresses and hopes to support this diversity through an interesting asymmetrical form.

Design intention 6: Open and inviting to complement existing pedestrian flow

Audits noted that both sides of the street are heavily used, that the Lygon Court Woolworths entrance was a key feature movement corridor and all along the median was used a lot as a crossing point. A multi-directional design with many ‘faces’ invites users approaching from different directions and complements existing movement paths to feel organic.

Design intention 7: Support safe transport, transitions and waiting

A mix of transport modes were observed including cars, taxis, rideshare vehicles, scooters, bikes, prams, skateboards and e-scooters. With cars entering covered parking, Ubers and rideshares dropping-off and picking-up and bikes and scooters being locked or parked the space felt like a transport hub. Very few pedestrians waited for traffic lights to indicate green for crossing. The design faces out so that anyone dwelling might be aware of surrounding traffic flow. An urban dance floor or handball court, for instance, would not have been a good fit for the dynamic nature of the space.

The following design principles also guided designers.

• Comfort: The design includes comfortable seating surfaces, giving people the option of an armrest or open seat and includes wooden slanted backrests for comfortable dwelling. Materials were chosen for comfort, with steel as the structural material and timber for contactable areas.
• Accessibility: The surrounding slab under the chair was designed with a larger circulation space on the side with an accessible ramp nearby, to accommodate a wheelchair user. Someone with a walker, pram or trolley might also find this space useful.
• Enhance the character and create a sense of place: The chair is natural timber with powder coated steel in 'Leaf Green' to reference the existing urban environment. An artistic form could improve the usefulness of the space while adding to its character - “meet you at the wavy chair!”.

The design is currently being manufactured in Adelaide by Rowlands Metalworks, overseen by designers at Spark Furniture. See a snapshot of the manufacture process below. It will be installed in December 2022.

Data will be collected in a variety of ways over a minimum of six months to understand the impact of the design.

One way of capturing data will be undertaking the same place audits that we carried out in November 2021. This will help us understand the impacts of the design on the space through observing and talking with users as well as compare activity before and after the chair was installed. For instance, we might learn that the chair impacts user paths, time spent here, the ways users occupy other areas of the median, social activities, diversity of the users or perceptions of safety.

We will invite the community to share feedback to measure if the design supports the desires and needs expressed by users in the place audits; are there more places to sit, does the chair make the space more efficient and useful, is the space more comfortable, does it inspire more activation?

We’ll also be using sensors to collect quantitative data. This will aid our understanding of how the chair is being used. For instance, we might find that it's heavily used at certain times of the day, that some parts of the chair are more popular than others or that there are trends in relation to weather conditions and chair usage. We’ll publish a data dashboard to share insights with the community.

Findings from this stage will inform our approach to the design of our city spaces and infrastructure by shaping when and how we use data.

• Digital Matter Yabby Edge - accelerometer
• VivaCity Artificial Intelligence Road Survey sensor (AIRS) - pedestrian and bike sensor

• Usage of the existing bench in the median strip
Most popular times of use for the new design
How people are moving around the new design
Popular modes of transport in this area
Key movement pathways
Speed of movement of bikes and pedestrians at various times.

Data from the Yabby Edge sensor on the existing bench is collected to help us understand how people use this seat. This will help us to understand how people use the existing chair in comparison to the Social Spaces design which will inform future planning and design. These sensors are also used in Argyle Square, Royal Park and at Micro-Labs. You can see some of the data captured by this type of sensor on our public dashboards for Micro-Labs and Argyle Square.

We will also collect data using an AIRS sensor to understand the flow of people around the design and their transport modes. For instance, it will be able to tell us the busiest times, most popular modes of transport and the kinds of movement around the design. We have also installed this type of sensor at Southbank Promenade to learn more about how people use different modes of transport in this busy city thoroughfare.

Data will also be collected via observational audits in 2023.

Yabby Edge sensors use low power wide area networks or LoRaWAN to transmit the data. The VivaCity AIRS sensor uses 4G networks to send data. The data is sent to City of Melbourne’s data lake in AWS Australia.

The data is used by City of Melbourne to measure the impacts of the Social Spaces design and inform how we plan and design other outdoor spaces like this. The data will be shared with the community via a dashboard on this page soon.

This data will be stored permanently for any possible future analysis.

For the AIRS sensor, no video footage is stored or transmitted. The only data that is stored is created from analysis of the footage.