As governments across the globe continue to shift toward the concept of smart cities, there is an extremely strong demand for solutions that can alleviate the risks associated with environmental disasters.
With the increasing number of cities and people moving to urban areas, there is an unprecedented rise in the risks associated with natural calamities. The series of hurricanes that recently battered the U.S. and neighboring countries is an example. As governments across the globe continue to shift toward the concept of smart cities, there is an extremely strong demand for solutions that can alleviate the risks associated with environmental disasters.
There are several methods of predicting environmental risk at the moment, but these fail to provide a comprehensive picture of issues and therefore fails in helping initiate necessary precautions. However, a Boston, Massachusetts-based company has a better idea.
Palmos is an environmental risk monitoring startup that designs and builds innovative multi-sensors to enable its analytics. The Palmos system uses low-cost, on-the-ground sensors in conjunction with open source data and physical modeling to produce more complete real-time risk maps of an area.
“Our primary target markets include cities and natural catastrophe reinsurers,” said Gregory Kollmer, CEO of the company. “City governments benefit from the Palmos system for early warnings in the short-term and better urban planning for improved risk mitigation in the long-term. Reinsurers benefit from improved risk assessment of insured assets to create more accurate pricing models and significantly improved real-time damage assessment for use in innovative parametric insurance applications.”
Behind the solution
Palmos’ system is composed of a high-density configuration of solar-powered MEMS sensors. Each node transmits real-time environmental data to a relay node via RF technology. Each sensor node has an average durability of three to five years without direct intervention.
“The data from the sensors are pre-processed in the sensor node and securely transmitted to the relay node,” Kollmer said. “The relay node sends the processed information to our cloud servers via a cellular network. In the cloud, the data is further processed and then fused with open-source meteorological data and input into physical modeling. The resulting analytics are then pushed to the user. The user interacts with the Palmos data dashboard through an online web portal or through a phone application, where a risk score is displayed in the form of a ‘heat map’.”
He added that the current methods of predicting environmental risks in any given area are limited by the severe lack of local, continuous data. The dominant methods include some combination of general meteorological data along with intermittent satellite data for more advanced cities.
“For less advanced cities, environmental risk monitoring is severely limited and often non-existent,” Kollmer. “Palmos uses specially designed lower cost sensors to have the local, real-time data needed to make truly local risk and damage assessments. This technology is only plausible due to very recent trends like improved microsensor manufacturing, cloud computing, LPWAN technology improvement, and significantly improved machine learning methods.”
Urbanization and the rise of AI
Needless to say, the primary driver of demand for Palmos is the rise in the number of cities. But adding to this are concerns about climate change and how they are increasing the severity of weather-related disasters.
“The primary driver is increased urbanization in cities that are at increased risk of environmental risks due to proximity to the coasts. This is further exacerbated by the increasing unpredictability and severity of weather events due to climate change,” Kollmer said.
Technological developments are further expected to help the solution. Kollmer agrees that the rise of artificial intelligence and technologies like machine learning will have a major impact on their product and the industry itself.
“We are already using machine learning classification algorithms,” he said. “To go in more detail, the sensors we build are only vehicles to transform signals. They transform the information of interest (e.g. physical signals) into information that can be readily analyzed (e.g. digital signal). The analysis of that information through methods like ML or deep learning is the truly valuable system component because it produces comprehensible insights that are used to save money but more importantly human lives.”
The solution is still in the development phase but Palmos has already received positive feedback and support from governments in Brazil and the U.S. along with reinsurers based in Switzerland. The company has a city pilot launch planned in Brazil for 2018 with expansion plans.