NDS Cognitive Labs provided a solution to a water services company that enables the tracking of school and park drinking fountain conditions via a Sigfox connection to determine when water filters should be reordered.
Internet of Things (IoT) technology is helping school children in Mexico City access healthy drinking water, with cloud-based software managing the pressure, consumption and conditions of water at each drinking fountain. NDS Cognitive Labs, an artificial intelligence (AI) and tech services firm, has partnered with a Latin American water services company to develop and deploy the IoT system.
The solution consists of wireless sensors on each fountain to capture water-pressure data, as well as NDS Cognitive Labs software that manages that data, along with information from multiple local authorities regarding potential water contamination for each site, so that the water company can visit specific fountains when necessary. The system gathers the sensor data via Sigfox-based IoT or a cellular network. The NDS software then analyzes the collected information and populates a dashboard with a map that allows users to obtain real-time readings regarding water quality, velocity, pressure and usage.
The Greater Mexico City metropolitan area has 19 million inhabitants living at an elevation of more than 7,000 feet, many of whom face water shortages. The State of Mexico provides bulk water from Comisión Nacional del Agua (Conagua), which is managed by a variety of local agencies and municipalities. To transport that water, the nation has nearly 11,000 kilometers (6,835 miles) of distribution lines, with several million water connections, some of them illegal.
NDS Cognitive Labs, based in the city, was launched 15 years ago to provide software-based solutions for smart buildings and smart agriculture (for instance, a system to track the production of dairy products), according to Gustavo Parés, the company’s CEO. The firm began working with GE Digital in 2015 to develop Industry 4.0 manufacturing solutions.
The company’s software provides AI and machine learning for sensor-based data that can be used by businesses’ operational teams. “We wanted to take advantage of how unstructured data supports structured data for decision making,” Parés says. In most cases, manufacturers face challenges related to disparate systems used in manufacturing, with a separate IT system that lacks access to the manufacturing data.
In 2019, NDS Cognitive Labs partnered with a water services company that provides infrastructure which collects and moves potable water throughout Mexico, including cisterns, piping and drinking fountains. The company had recently won a government contract from the State of Mexico to extend a source of clean, drinkable water to schools and parks. The goal, Parés says, was to ensure children had access to running, high-quality water in the places where they studied and played.
Simply deploying drinking fountains, without intelligence technology, would not solve drinking-water problems in the long term, Parés explains, because it would be too challenging for the water company to send employees to regularly visit each of the hundreds of fountains to track water pressure and determine whether filters needed to be replaced. Additionally, multiple entities, such as local water authorities, provided inspections at some sites, but the water company lacked access to those reports. Therefore, NDS Cognitive Labs began working with the firm to set up a single, unified system network.
“The first stage we had involved mapping all the areas [where the fountains were installed],” Parés says, “to make sure we could make a centralized decision-making platform.” The system had to be able to consume information accessed from a variety of public authorities. “The second challenge was how we could make sure we have enough filters to address water conditions.” Some fountains may require filter changes more often than others, based on the condition of the water being filtered or the use rate by people onsite.
The resulting system consists of a blend of manually recorded data and IoT-based filter information to identify filter replacement times. Each drinking fountain comes with an IoT sensor device that measures water pressure, volume and consumption. The fountain’s pressure data is then analyzed to determine whether a new filter is required. An increase in pressure could indicate the need for a filter replacement, while the heavy use of water could mean the same thing.
The system has a unique ID number linked to that fountain in the software. Data is periodically transmitted to NDS’s server via cellular or Sigfox connectivity, depending on a school’s location. Sigfox, a French global network operator, provides IoT connectivity using low-power wide-area transmission. If the fountain requires a new filter, the system can automate the ordering of a filter for that site. Additionally, the software manages data related to the water’s condition based on local inspections, and the water or health authorities share that information with the automated system.
Based on that information, NDS Cognitive Labs’ solution can prompt the scheduling of service personnel to visit a specified fountain. The software comes with AI functionality to calculate conditions for a fountain that might not have been inspected, based on its proximity to other fountains that have been inspected. “We take data from different sources and apply common ranges,” Pares says. This mathematical interpolation method, known as Kriging, estimates a variable depending on the geographical location.
The sensor data enables the water company to set acceptable thresholds based on pressure or volume, for instance. While the software could generate 100 reports from the data being captured, Parés says, “We had to focus on just two or three indicators.” The system has been deployed across Central Mexico, including in Mexico City. In the long term, it could be extended throughout the country, though the government has placed a pause on the deployment for administrative and political reasons.
The partnership between NDS Cognitive Labs and the water company has enabled a network of clean water that could serve thousands of children, with limited human intervention. “It has been a good first approach,” Parés states. “The water company is the water experts, but they don’t have the resources to collect and manage that information.”