Climate Change and IoT: How to Save the World

The use of advanced data analytics, powered by AI, is helping humanity to address climate change in an intelligent way. That is the reason why so many technologists and business leaders, around the world, are now partnering to reduce the emission of dangerous greenhouse gases.

The entire effort is based on the harsh realities brought home to everyone by the landmark Paris Climate Agreement of 2015. That treaty, signed by more than 188 national governments, aims to limit the global temperature increase during this century to 2 degrees Celsius above preindustrial levels, with a stretch objective of only 1.5 degrees. According to the consortium of Nobel Prize winners at the Intergovernmental Panel on Climate Change (IPCC), both of those scenarios will require “rapid, far-reaching and unprecedented changes in all aspects of society.” To understand technology’s increasingly central role in the unfolding climate change drama, and for better insights into how tech might save the world, we asked four world-class experts to share some of their best insights.

Climate change will continue to be the defining challenge and opportunity for the rest of the 21st century.
_{Alex Mitchell, Los Angeles Cleantech Incubator}

Alex Mitchell is Senior Vice President of Market Transformation at Los Angeles Cleantech Incubator (LACI). His work starts from a simple premise: “Climate change will continue to be the defining challenge and opportunity for the rest of the 21st century”. Reaching the targets, as agreed by Paris Climate Agreement signatory countries, will require an all-hands-on-deck mentality for policy makers, businesses, and technologists. Mitchell thinks that “advanced data will play a vital role in helping humanity address climate change in several key domains. In particular, our transportation and energy systems, which represent more than half of America’s greenhouse gas emissions, are ripe for analytics-based breakthroughs. Three in particular stand out: real-time multimodal commute data, improving traffic flows at traffic lights, and real-time energy consumption data”. Transportation accounts for 28 per-cent of US greenhouse gas emissions, with passenger cars accounting for more than half of that total. Unsurprisingly, reducing private car usage is one of the holy grails of addressing climate change. Mitchell points out that “LACI’s public-private Transportation Electrification Partnership, for example, calls for a shifting of over 20 percent of all trips in single-occupancy vehicles in Los Angeles County to zero-emissions public and active transit by 2028”.

All in It Together: The concept of ecosystem includes the actors that form part of it, connected to each other through various relationships. In addition, it is necessary to define the limits of that ecosystem, which can be established by the evaluation range of the product/service system of consumers or the perception of those consumers (Source: Los Angeles Cleantech Incubator).

However, getting people out of private car ownership will require robust commute reliability information. Today, nearly every form of transportation, beyond the private car, suffers from low-quality real-time data, from inaccurate bus arrival information to incorrect scooter locations. Multimodal trip planners, such as the Transit app, do something that Mitchell finds useful: “Combining data sets across scooters, bikes, buses, ride hail, and walking to help travelers make better informed decisions about how to best get from point A to point B”. Even major public transit operators, such as Los Angeles Metro, are adopting such platforms for their own consumer interfaces.

Despite the proliferation of apps – such as Transit, Moovit, and Google Maps – these companies are still in their infancy. They are only beginning to become proficient at harnessing advanced data to help improve multimodal travel. As Mitchell points out, “the goal is to discourage private car usage”. For example, few, if any, of the apps allow for robust consumer input preferences, such as the ability to signal a longer tolerance for walking. Mitchell notes that these apps aren’t yet handling “robust multimodal execution, such as ‘your commute will now be 15 minutes shorter if you exit the bus in two stops and take bike share”.

Daylight Simulation: At cove.tool, software developers are fighting climate change by helping interested parties use data-driven design through automation and cost optimization (Source: cove.tool).

Mitchell is quick to turn attention to what he calls “the plebian traffic light,” which is now over one hundred years old. “It hasn’t changed much over time. While it does reduce automotive collision fatalities, it hasn’t kept up with our climate challenges”. Few traffic lights rely on any real-time data of city-level traffic, such as would be accomplished by adaptive traffic control. Mitchell thinks they should assess and analyze “individual vehicle-level data sets to determine when the light changes from red to green”. Instead, they often rely on unsophisticated, preprogrammed instructions. As a result, an SUV might sit idling at a red light for two minutes, while the road with the green light may have no vehicles at all. The net result is a loss of travel time and unnecessary greenhouse gas emissions.

We know that algorithms can’t make adaptive changes on their own.
_{Micah Kotch, BMW URBAN-X}

Dozens of companies can be cited as actors in this transport revolution, including these: XTelligent, one of the companies embedded inside LACI; Rapid Flow Technologies, and their SURTRAC solution; and Siemens. According to Mitchell “all of these, and many more, are aiming to handle the massive IoT-generated data sets from cars, trucks, buses, bikes, and cell phones to orchestrate traffic flows in cities, both at the system level and at the individual light level”. Shifting the focus from mobility to real-time energy consumption, Mitchell is thinking about the abundance of data in this realm: “Few consumers think about their energy consumption until they get their electricity and gas bill. So they fail to realize that tiny changes in hour-by-hour increments have huge impacts on the energy generation system”. When air conditioners go on, for example, a natural-gas-fueled “peaker plant” may be forced into service to cope with the incremental demand, supplementing the base energy load generated by renewables such as solar and wind.

The End of Potholes: RoadBotics has developed technology that identifies potholes, alerting authorities through an interface that zooms into the tiniest detail of the asphalt beneath your wheels (Source: RoadBotics, Inc).

Where the system fails today is in alerting consumers, in real time, of the consequences of their behavior and giving them the choice to modify accordingly. Mitchell cites, as an example, the work of one player: “Ohm-Connect Corp. gamifies consumers’ engagement with their home energy consumption, saving an average of US$100 on their bill and paving the way toward a more renewables-based energy system. As IoT devices in the home proliferate, and as smart meters scale, expect a growing number of ways to help consumers make the right financial and environmental decisions”.

The second leader whom we turned to is Micah Kotch, the Managing Director of the BMW Group’s URBAN-X. From his base in New York City, Kotch shines the spotlight on a few of the most innovative solutions that have recently been developed by URBAN-X teams. During these challenging and adaptive times, Kotch worries that it’s easy to take our eyes off the goal of climate change adaptation and mitigation. From the point of view of the URBAN-X teams, “building resilience in our cities is a generational challenge, and technology is a critical tool in our arsenal. At URBAN-X, we believe that by investing in and supporting entrepreneurs using data, machine learning (ML), and AI to address society-scale urban issues, we can build iconic companies for the 21st century”. Kotch thinks that “while we’re in uncharted territory, from past experience we know that algorithms can’t make adaptive changes on their own”. According to research firm Carbon Brief, global carbon emissions could fall by around 2 billion metric tons this year, equivalent to 5.5 percent of last year’s record emissions. That would represent the biggest drop since World War II. However UN projections say that holding global temperature rise below 1.5°C will require even greater annual emission reductions of 7.6 percent over the next decade. For Kotch the requirement is fairly straight forward: “We need entrepreneurial heroes who challenge the status quo, use data to improve the lives of real human beings, and have the tenacity and resolve to meet this most critical of mandates”.

As with all technology, the tool itself is not a silver bullet.
_{Seth Robinson, CompTIA}

Start-ups are outliers in solving difficult problems, quickly. URBAN-X has built a portfolio of 51 companies reimagining infrastructure; roadways, building design, construction and operations, and wastewater systems. Kotch is justifiably pleased with the results: “From electric vehicle charging, to drone transportation, to the cold chain that our 21st century food system relies on, our teams are harnessing ML to improve inefficient processes, reducing time, cost, and emissions around the world”. URBAN-X portfolio companies using advanced data analytics to address our climate emergency include these four young companies, each deploying breakthrough technologies:

• Cove.tool is providing streamlined automated analysis that helps architects, engineers, and contractors achieve energy, daylight, glare, water, and carbon targets while reducing construction costs. Teams reduce analysis time by 66 percent per project by using automated performance modeling. Kotch has this to say about the company’s key idea: “As building energy codes are becoming stringent across the world, designers can help cut construction costs by 2 to 3 percent to meet these new codes. Cove.tool’s ML algorithms provide visibility into thousands of alternatives. Realestate developers can invest with confidence while reducing construction costs to meet energy regulations and tenant needs”.
• RoadBotics uses advanced AI and computer vision to produce detailed maps of a government’s road infrastructure problems. These issues include things like potholes and utility patches, but what they really help communities to unlock are effective preventative maintenance strategies. Kotch explains: “So just like you perform maintenance on your car to help it last longer and prevent major fixes later, the same is true of roads. More than 200 governments around the world use RoadBotics AI to reduce the use of asphalt and concrete, all while achieving better road infrastructure for everyone”.
• Sapient provides a turnkey plug load management solution that couples machine learning, an enterprise SaaS web application, and a comprehensive deployment of smart plugs and smart power strips for commercial buildings. The Sapient web application collects energy data through smart devices and controls power delivery at each individual socket for every plugged-in piece of equipment throughout a building. Sapient’s web-based analytics platform leverages AI to do something that Kotch thinks is important: “generate insights regarding energy inefficiencies, space and equipment utilization, safety, and equipment performance. The result is significant reductions in energy consumption, electronic waste, and carbon emissions”.
• Hades uses ML to automatically detect defects in sewer inspection footage, predicts how defects evolve over time, and identifies ideal repairs. More than 5 million miles of sewer drain waste and storm water from cities around the world, which is critical for environmental protection and public health, especially during periods of heavy rainfall amplified by climate change. Extending sewers’ useful service lives not only saves money but also prevents significant greenhouse gas emissions caused by excavation-intense construction.

Where Did Winter Go? Average December temperatures in the U.S. are more like March or April, meteorologists say. In Kansas City, average temperatures in January 2020 were 12 degrees above normal, an indicator of increasing climate change (Source ©: PRISM Climate Group | Oregon State University).

Both coronavirus and climate change require rapid responses For Kotch and the team at URBAN-X, “the faster we act, the less disruption and loss of life. In both cases, the need for local action, creative solutions that leverage data, sensors and analytics, and good public policy based on science has never been clearer”. We also know that “AI” has become an overused buzzword in its own right. Kotch thinks that “in order to get sustained emissions reductions, technology needs to work alongside policy and markets”.

The third leader providing us with his insights is Seth Robinson, the Washington DC-based Senior Director of Technology Analysis at Comp-TIA, the world’s leading tech association. CompTIA’s Board of Directors includes C-Suite executives from some of the world’s largest companies, including Hewlett Packard Enterprises, SNC Lavalin, EY, Comcast, SAP, and Cisco. Robinson thinks that one big “part of the appeal in emerging technologies is the ability to leverage computing power that has rarely been accessible”. That power can then be applied to extremely complicated problems. A perfect example is the way that artificial intelligence can be used to attack the issue of climate change. Whereas earlier software programs were deterministic, taking in defined inputs and producing repeatable results, modern AI programs act more on principles of probability. Massive data sets with varying amounts of structure can be fed in, and the algorithms look for patterns and correlations in order to produce results that have a high likelihood of being correct. This type of activity is a perfect match for climate analysis, which has always been a guessing game using lots of data. Robinson’s research shows that there are three areas “where AI can have an impact in trying to solve climate change”.

Smart Socket: Sapient couples machine learning and an enterprise SaaS web application with a comprehensive deployment of smart plugs and smart power strips (source ©: Sapient Industries, Inc).

• First: AI can assist with the ongoing work in predicting weather patterns and events. Climate modeling got its start in the 1960s and has been constantly evolving since then. One of the more recent updates has been the addition of data science principles, commonly known as climate informatics. AI algorithms are a natural next step in this activity, driving the ability to reconstruct past climate events and to improve the prediction of the future, especially concerning extreme events.
• Second: AI has a major role in predicting the fallout of those extreme events. While rising sea levels and intense storms may be accepted as likely outcomes of climate change, it can be hard to visualize the actual impact. Using data from previous events and models for the future, Robinson concludes that “AI can build simulations showing the potential scope of damage, which provides decision makers with tangible information as they form their plans”.
• Third: AI can be applied to the gargantuan task of reducing carbon emissions. This may be the most diverse field in the intersection of AI and climate efforts. The type of activities can range from carbon tracking across a geographic region to optimizing electricity usage or providing suggestions for physical building operations. The findings from these activities can help individual businesses improve their carbon footprint and also help government officials create policies and regulations. For example, The Edge, a unique office building in downtown Amsterdam, is considered one of the most environmentally friendly buildings on the planet. It uses an AI system which talks to 30,000 sensors to control lighting based on activity – all with the aim of reducing energy usage and moving toward zero climate impact. According to Robinson, “the system is estimated to save over $100,000 per year in energy costs. Climate change is poised to be one of the great challenges for the global population in the coming decades. Artificial intelligence can provide novel insights and automation, but remember that the results are still based on probability. As with all technology, the tool itself is not a silver bullet. It is critical to also have the right expertise in place to interpret the data and build guidance based on the results”.

The fourth leader we’ve turned to is Steve Westly, a famed Silicon Valley investor who served for years on Tesla’s Board of Directors. For decades he’s been a top-tier tech leader and venture capitalist, with multiple mega-successes to his name, a journey that started long ago in his role as the first CFO of eBay.

ML and AI have the power to dramatically change the world of the future.
_{Steve Westly, Silicon Valley Investor}

Westly starts from the core premise: ML and AI are helping humanity address climate change “by aggregating and analyzing disconnected data sources to create more accurate predictions, increase energy efficiency, and optimize current systems”. He thinks that, as we continue to rely on renewable energy sources, “utilities and energy providers need more accurate methods of predicting energy consumption, in both real time and the long term”. ML and AI play a role in using data sources such as local weather, climate patterns, and household consumption behaviors to help predict future demand. This will lead to more effective energy distribution, which is pivotal in minimizing the carbon footprint.

In the light of Westly’s investment experience, one key aspect of reducing CO2 emissions is the electrification of vehicles and fleets, as electric vehicles (EVs) have a 54 percent smaller carbon footprint compared to gas vehicles. “At the individual consumer level, ML and AI algorithms can improve battery energy management to increase mileage of each charge and reduce barriers for widespread adoption of EVs”. By 2030, it’s predicted that electric cars will account for 70 percent of vehicles sold in China, 50 percent in Europe, and 30 percent of all sales in the US and Canada, leading to more energy demand. Vast new systems must be “designed and built to meet this influx of demand, from public fast charging networks to embedded control schemes and grid upgrades”. Westly points to Weave Grid Corp. as “a prime example of how companies are utilizing ML/AI to provide a technology advantage, helping utility companies make better decisions and manage this energy efficiently, while supporting continued reliability and meeting key cost-effectiveness goals”.

Submeter Nation: There is a growing wave of regulation in the United States focusing on the metering of electrical consumption. Cities and states are establishing ever-tightening directives to reduce energy use and align payment with consumption (source ©: Aquicore, Inc).

Integrating both historical and real time data from a diverse array of sources, Weave Grid aims to provide cutting-edge monitoring, prediction, and optimization tools that will enable and accelerate the multi-industry effort to electrify transportation. Wasted energy in buildings is another environmental issue that has been changed by ML and AI. Commercial and residential buildings account for about 34 percent of greenhouse gas emissions in the US (in New York City, 75 percent of the city’s carbon footprint comes from building emissions).

Smarter Cities: Whether smart building technologies, transportation, public health, public safety, or smart infrastructure, there are plenty of ways for solution providers to find a market to build a successful practice (Source ©: CompTIA.org).

It’s now estimated that, in the US, 30 percent of that energy is wasted. This issue has been mitigated due to widespread adoption of networked and highly sophisticated energy meters that provide real-time data every 15 minutes (vs. old models once a month). Westly is impressed by the fact that “this access to Big Data has enabled companies, like Aquicore Corp., to use ML and AI to build intelligent software to drive actionable insights through software that collects, analyzes, and executes actions to improve energy efficiency within a building”. This sort of “smart” monitoring system allow for reductions in energy waste and CO² emissions by enacting automatic actions based on algorithms and insights from the data monitoring system. Additionally, a smart building could also communicate directly with the grid to reduce the amount of power it is using if there’s a scar-city of low-carbon electricity supply.

Westly, the former State Controller of the State of California, sums it all up very well: “ML and AI have the power to dramatically change the world of the future – and perhaps nowhere will it be more important than in the battle against global warming”.

READ MORE ARTICLES