Using AI & IoT to Track & Understand Visitor Flow

Managing visitor flow with AI and IoT

When we talk about IoT we often talk about the power of data-driven decision-making. MeltSpot, the innovative Swedish tech company, has built one of the most powerful tools available for real-world data analytics, which can bring your decision making to a whole new level. How do they do this? By installing sensors that track visitor flow and turning the gathered data collected into actionable insights. These sensors are installed in places like shopping malls, concert halls, airports or around smart cities at every crosswalk and entryway – basically anywhere people are moving. The sensors are a ‘mobile nest’ in miniature format, but instead of communicating, MeltSpot sensors ‘listen’ on the same frequencies phones transmit data over, allowing MeltSpot to count and track devices and create aggregated statistics based on movement and visitor flow, all while being fully compliant with European Union GDPR regulations. This includes KPIs such as the number of people, but also how long they stay in a particular spot and how often the average visitor is recurring. And this information can be invaluable to all kinds of stakeholders.

To put it another way, the MeltSpot platform and solution gives customers hard data on visitor flow and that hard data helps businesses plan for a host of things, including infrastructural activities and changes, as well as future investments and measuring the impact of solutions – and all this leads to better outcomes and better customer satisfaction, while the privacy of individuals is also safeguarded.

“A smart city might use our sensors to plan roadworks or to understand that certain areas of a park are not properly lit because they can ‘see’ that people take a detour during the evenings,” explains Alexander Bucur, CEO MeltSpot. “A shopping mall might be interested in which stores pull in the most customers and at what time, which will help them when planning for the next tenant or setting the rent. All of it is data-driven decision making, and by that, I mean it helps with planning both immediate decisions and long-term ones. This will become increasingly important to cities that need of critical data that will allow them to continuously improve the environment and experiences for their visitors and residents . This kind of data will also greatly assist in urban planning, helping cities cope with increased populations and plan better for the future.”

The MeltSpot solution gives real facts backed by hard data – and with those real facts you can better plan for your infrastructural activities and changes, as well as future investments. You can also use the same tools to measure the impact after you’ve implemented your solution, making adjustments based again on data and leading to better outcomes and for businesses, increased conversation rates.

MeltSpot is also used to assess customer satisfaction in the form of crowd analytics, meaning where are the crowds and bottlenecks in the area? Where are the queues?

Marcus Malmqvist Co-founder MeltSpot

“Obviously, during the pandemic MeltSpot was a great tool to see where crowds were gathering, allowing facility managers to address those challenges quickly and safely. Even now, though, as things have gone somewhat back to normal, queues and crowds are normally not something that gives a good experience, so better understanding, response, and planning can really be a benefit.

“When we’re talking about events where large crowds gather, such as concerts and sporting events, MeltSpot can be utilized both in the moment to understand crowd behavior, but also in planning for future events, such as there aren’t enough food and drink stalls to handle demand, or certain bathrooms have unusually long lines. Addressing these challenges will greatly improve customer satisfaction, because let’s face it, few people like standing in long lines. So, we offer real time reaction, but also forecasting, for future planning capabilities.”

The importance of privacy

When it comes to data, privacy is imperative and MeltSpot takes privacy extremely seriously. The MeltSpot team has worked extensively on ensuring that all data they collect is in accordance with GDPR, as well as Swedish privacy laws. For MeltSpot, one phone is one person in terms of data, but they never actually ‘see’ who the person is.

“The entire technology stack, from the point of data collection to end-user visualization, relies on aggregating data into KPIs and percentages, so we’re never tracking an individual phone,” explains Marcus Malmqvist. “We’ve also built everything around ‘privacy by design’, which basically means that not even our team, who have full access to the database, can find even our own phones in the system. This is because everything is anonymized on the way in, so essentially, the sensors only pick up a few percentages of data from a phone, and we only collect enough to identify relevant phones roaming in the area. This ensures that privacy is not compromised, and that people aren’t being individually ‘tracked’.”

MeltSpot also ensures that there is a certain amount of ‘collision’, so a certain number of phones will look identical in the system, which gives another layer of anonymity. So, if they were to get a request from the authorities or anyone else about a specific device, they can’t answer that because it could be 20 or 30 people in a space who the phone might belong to.

“We have multiple layers to protect people’s privacy because we know how important it is not just from a regulatory point of view but also from a ‘people’ point of view,” says Alexander Bucur. “On top of that, our customers get a dashboard with only aggregated data with percentages, such as ‘200 people came here, 40% were recurring, these people were interested in lunch’ – but we don’t know who is who, and neither do our customers.”

AI & IoT

The data MeltSpot collects can also be very noisy. Just by the way it is collected allows MeltSpot to see things like printers, people walking around with an iPad and an iPhone, and both are connected with a SIM card and walking together.

“By leveraging AI models we can make exceptions that dynamically pinpoint and filter who is a customer and what is a connected smart camera or other IoT devices,” says Marcus Malmqvist. “We set up hard parameters and rules that interpret the data, which gives another level of quality because if we hand it over raw it’s too much data to sift through and analyze. We also do anomaly detection, which could be people behaving a certain way.”

So, why cellular connectivity as opposed to Wifi or other connectivity options? For MeltSpot, other solutions weren’t right for the amount of data being collected and the required reliability of the real-time connectivity to MeltSpot sensors. Having full end-to-end control of the flow of data and not relying on on-site, third-party networks, allows for a more scalable and reliable solution.

“Our entire sensor is an IoT device with a SIM card in it,” explains Alexander Bucur. “We used to use the customer’s WiFi but it was too unreliable, and a big hassle with passwords being changed and other things. When you’re working with this many sensors in one place you want them all to be aligned, otherwise you can’t really visualize the flow of people, so moving to incorporating SIMs and stable cellular connectivity was the natural next step.“

Why Tele2 IoT

For a fast-growing company working on the cutting edge of technology, it’s important to work with the right partners who can help you move fast and give you the kind of support you need to succeed.

“Working with Tele2 IoT has been great,” says Alexander Bucur. “We were really impressed by the onboarding in particular. We got in touch with two people who helped us get started really quickly – so from us calling Tele2 IoT and asking for information to being up and running was less than five days total, so very, very quick. We had the SIM cards the day after we ordered them, we tested the first one the day after that, and it was very smooth. Also, 2CONTROL (Cisco IoT Control Center) works really well for us as well and integrates into our existing setup well.

The future

When looking to the future, MeltSpot has great ambitions. They have developed a very strong and unique tool, one that is much more complete and accurate than most of their competitors, so they are advancing and accelerating as quickly as possible.

“We are not done developing,” says Marcus Malmqvist. “We are working on securing IPs on some of the technology we have developed, and we have designed the MeltSpot solution to work everywhere. It is business market agnostic, so we can work with any type of customer where people flow, and movement is involved. Different customers have different needs and uses for the technology, so we keep that in mind when developing.”

While MeltSpot is initially focusing on several specific markets, their solution is global and expansion means they can work with everything from airports and train stations to amusement parks or smart cities.

“Only the imagination limits us – and we’ve got plenty of imagination,” laughs Alexander Bucur. “We’re building our foundation and part of our strategy is to work with high level customers who understand the needs of this kind of data-driven analytics, which means we’re also looking into partnerships with important customers and distrubutors who work with this kind of multi-level data. We’re not just ‘people counters’ for stores, we want to be in a game where we make a difference for important decisions and bring real value to our customers.”

If you would like to learn more about how IoT can enable your business, please get in touch.

Using IoT and AI to Enhance Workplace Safety

Customer case

The Tele2 IoT Smart City White Paper

White paper

Seven Ways IoT Enables Retail

Customer case