Wits University innovators have developed a novel Wi-Fi-based solution for managing and monitoring foot count, social distancing and consumer behaviour in shopping malls. The solution uses artificial intelligence (AI) to allow retail managers to make future predictions, improve security and deliver intelligence to stores about shopper behaviour.
Its inventor, Dominique Adams has been awarded R100 000 worth of support from WITS Enterprise towards developing the business so as to take the innovation to the next stage. This award was for the winning pitch of the latest Prospector@WITS course run by Wits Enterprise. Development of the Wi-Fi-based, AI-enabled solution began in 2020, shortly after the COVID-19 pandemic emerged in South Africa, triggering the concept for the social distancing tracking tool.
Adams explains; “Our innovation stemmed from an existing product for Wi-Fi signals to monitor social distancing in public spaces. We engaged with stakeholders in the retail space to determine whether there was a need for a solution that, with the application of AI technology, would give them valuable insights into consumer behaviour and help them make predictions.
“Combining Wi-Fi technology with AI, our tool allows mall managers to monitor the complex behaviour of people in real time as well as predict their future behaviour. This gives stores data-based intelligence while providing mall management teams with an additional tool for enhancing security. Our tool also allows shopping malls to monetize from the deployment of WiFi, and importantly obtain intelligent foot count and other data for strategic planning and evidence-based decision-making.”
As an example, Adams says mall managers can measure the overall popularity of stores and see the length of time a shopper spends at a particular store. The data gathered from these measurements are valuable for establishing shopper behaviour and for informed decision-making and planning.
Adams’ DataConvergance team is developing the solution under the leadership of Professor Bruce Mellado from WITS University’s School of Physics. The team, consisting of data scientists and AI specialists, includes Xifeng Ruan, Kentaro Hyashi, Finn Stevenson and Benjamin Lieberman.
In developing the tool from scratch, they had to learn how to harness the large amount of data available from WiFi systems and solve challenges around Wi-Fi signals bouncing off walls, the distortion of positioning of people and the complexity of terrains. In adapting the tool for shopping mall application, Adams recalls the biggest challenge being reluctance of mall managers to assist with the market and needs analysis.
“It has been a huge team effort. Being awarded the seed funding is huge achievement for me as well as our team, a great reward for all of the long hours spent. The knowledge gained through the Prospector@Wits Course and the valuable guidance from my mentor, Dineo Masokoane, were critical contributions to the successful outcome of the project proposal.”
Commenting on the solution’s potential, Prof Mellado says: “Given the small number of businesses contemplating or already adopting AI, there is a huge opportunity for this solution to be successful in the retail environment. If the solution is successfully implemented, DataConvergence could be a trailblazer in the deployment of AI in the retail sector.”
Ela Romanowska, director of innovation support at Wits Enterprise, says that this is another example of WITS University’s response to providing solutions to societal needs, and is a worthy winner of the Prospectro@Wits end of course pitch session.
Dineo Masokoane, innovation support manager at Wits Enterprise, says: “The solution will enhance data collection by exploiting and deriving greater benefits from the Wi-Fi availability in malls. This approach enlarges the scope of information available to management and planners to monetise their investment in Wi-Fi as well as optimise in-store strategy development.”
The next steps for Adams and his team are to refine the prototype for the retail environment and develop a pilot in a real retail environment before full and large scale deployment in shopping malls.