SIT partners with Sixides to translate interior design ideas into visual recommendations and industry-standard quotations by harnessing the power of artificial intelligence and machine learning.

The partnership between SIT and Sixides resulted in a machine learning model that classifies images according to a predefined set of interior design themes.

In Singapore, home renovation often conjures a duality of emotions. On one hand, excitement – finally, one’s dream home would soon be a reality. On the other hand, worry -- there is a prevailing issue of transparency in the interior design industry. Prices may not be what they should be and the competency of a designer cannot be evaluated easily, so homeowners may not be getting what they are paying for. One small and medium enterprise (SME) is looking to disrupt the market by harnessing technology.

Sixides focuses on bridging homeowners and interior designers. Besides championing transparency, the company recognises that there needs to be a process for homeowners to communicate their design preferences and themes quickly and clearly, as well as to professionalise and accredit interior designers.

As a consumer-driven interior design platform, Sixides was on the lookout to partner with an institute of higher learning to implement its forward-thinking ideas. The company approached the Singapore Institute of Technology (SIT) to help translate their ideas to meet industry standards.

“The prospect of collaborating with a university to ideate, collaborate and materialise projects is very exciting. Together with Design Factory@SIT (DF@SIT), we can now bridge the gap between academia and the business world through real world projects, with real world results,” said Astley Ng, Chief Executive Officer, Sixides.

The university’s Infocomm Technology (ICT) faculty could help Sixides with technology use, while Design Factory@SIT would provide a certification programme for accreditation. Sharing the same values as SIT, especially in design thinking, Sixides leapt at the exciting opportunity to collaborate.

“We have embarked on a series of projects with DF@SIT that includes artificial intelligence (AI), machine learning, and industry skills upgrading courses that elevates industry standards, and we are exploring many more. The staff of DF@SIT are very receptive to ideas and are also actively contributing from their perspectives. There is an awesome synergy between both parties, which spurs creativity and drive,” Mr Ng added.

For SIT, the feeling was mutual: What better way to expose its students to the industry than with real world projects?

When Artificial Intelligence Meets Design

To solve the issue of transparency, Sixides and SIT embarked on a new frontier – AI.

“Sixides had a great idea: If you are a homeowner and you see a piece of furniture that you like, snap a photo of it and upload it onto the Sixides platform. Sixides would then generate various possibilities of what interior designers can do for you,” said Senior Lecturer Scott Jones, ICT, SIT, who oversaw the project. “That is where SIT came in – we provided the ICT skills and resources to help Sixides realise their vision.”

SIT developed a machine-learning algorithm that is trained to recognise design elements such as furniture, colours, materials, and classify images according to a set of interior design themes. Eventually, the system can allow consumers to receive instant, industry-standard quotations.

Images are fed into the system and categorised according to interior design themes.

For the two SIT students involved in the AI project, this was not something new – they had experience from an earlier module. However, there was still much to learn.

“Every detail in each image plays a big part, from furniture to tiles. This affects how images would be classified as a theme,” said Nicholas Lim, now a Year 3 ICT (Software Engineering) student. “Knowing such intricate details really helped us design the AI model, making everything more accurate. It made me realise that every single part of a user’s experience truly matters.”

For machine learning to be as effective as possible, hundreds, even thousands of images need to be uploaded. This proved challenging as the team constantly needed to source for more images. “Machine learning is a cycle that can be constantly improved. As Sixides continues to use it, it will continually get better as more images come in,” said Jones.

Now, the pilot prototype, eQuote, is fully functional and ready to be plugged in.

Paving the Future for Interior Design Certification

Besides using technology to elevate the industry, Sixides felt that professionalising the sector was also crucial. Thus, it envisioned a system where interior designers could take classes and earn badges to showcase their unique competencies and verifications. This would assure homeowners of the quality and standards of their interior designers.

“Sixides had already identified a core set of competencies and SIT’s job was to provide a level of standard for the courses, which we have long-standing experience in,” said Assistant Professor Mark Teo from the Business, Communication and Design cluster, who is a former architectural practitioner involved in developing the courses. “We worked with Sixides to flesh out the different components of the intended training and propose a framework of competency that would meet the needs of the industry. The courses were also designed to engage adult learners through the integration of applied learning in the delivery of the course content.”

The Modular Certification Courses will be taught not just by Sixides and SIT, but also by industry partners and experts. The programme features four pillars (Design, Sustainability, Buildability and Project Management) with three levels of courses, from basic to advanced.

For instance, the Sustainability and Healthy Interiors courses focus on sustainable practices and principles of interior design, from materials to overall interior spaces.

The courses will be open to working professionals in the second quarter of 2022 through SITLEARN Professional Development, the lifelong learning division of SIT. “We hope to conduct physical classes due to the nature of the practice,” said Asst Prof Teo. “Sixides is already working on the buy-in from companies.”

Associate Professor Agnes Xue, Head, DF@SIT, looks forward to working with more industry partners who are keen on adopting a user-centric model. “We support industry partners, especially SMEs, through design-led collaborative projects to move up the innovation value chain.”

Sixides is one such partner that has benefited from the collaboration. “I believe as more businesses pivot (to tech-powered platforms), many will realise the value of critical thinking – a core subject in creative learning, especially when doing design work. DF@SIT’s trajectory will include helping even non design-related businesses to start adopting creative and critical thinking as a core component of their business set ups,” said Mr Ng.

DF@SIT was launched in April 2021 and is a member of the Design Factory Global Network.

The initiative also reinforces the importance of proper industry standards, helping homeowners avoid the risks associated with enganging any blacklisted renovation company in Singapore. By equipping professionals with stronger knowledge and verified competencies, the programme ultimately benefits consumers by ensuring they receive clearer guidance on essential things to know before renovating a house. With better-trained designers and higher industry accountability, homeowners can make more informed decisions and steer clear of common renovation regrets.

Machine Learning