With the climate change crisis looming, many are betting on novel technological solutions such as carbon capture. We certainly need these sorts of solutions and as soon as possible. But if we want to start reducing emissions now, software may hold the key.

About 51 billion tons of greenhouse gases are added to the atmosphere every year, the bulk of which are in manufacturing and energy. Global producers of goods like cement, steel and plastic account for roughly a third of greenhouse gases (GHGs), while the electricity sector comprises a quarter. While some emissions sources are obvious, others result from indirect activity (scope 3), making them difficult to measure and, consequently, to reduce.

As researchers work to make carbon capture cheaper and more effective, it’s crucial to find ways to reduce emissions today. Software presents one such opportunity. Aided by software, processes can become as much as 10% more efficient, reducing the amount of carbon that continues to accumulate in the atmosphere.

Here are three ways businesses can implement software to reduce their environmental impact:

1. Eliminating waste

The production of consumer and industrial goods has never been optimized to minimize waste. Put another way, some amount of waste has always been justified to create products economically at scale. As a result, engineers reluctant to churn out products of inferior quality tend to add extra raw ingredients, over-estimating rather than under-estimating what’s actually needed.

Software can transform this process. Using machine learning, a piece of software can learn the complicated production process of an item in a matter of minutes, across a wide range of sectors: from a steel beam to a pint of ice cream. Such software can then figure out how to produce the same item with the minimum necessary amount of waste — immediately reducing carbon emissions.

2. Optimizing energy consumption

In the energy sector, software has already become a major driver of efficiency. Once a small area of investments, software now serves as a means of controlling the entire energy process, from generation and storage to distribution and consumption. Last year, the power sector spent an estimated $3 billion on software to optimize the performance and costs of generation and grid assets.

Other industries that rely heavily on fuel and electricity are following suit. AI software is being used to optimize logistics and trucking pathways. In freight shipping, many companies are adopting machine learning (ML)-based route optimization. Such software can design routes based on the most efficient pathways, reducing fuel consumption and helping shippers maximize profit.

Traditional manufacturing also relies on methods of production that cause high percentages of emissions, such as burning coal and fossil fuels. Software can optimize for energy usage and facilitate exploring more energy-efficient ways of producing. Without software, this would be costly, requiring time and a significant amount of physical experimentation. With software, we can shorten the experimental duration and latch onto lower-energy approaches to manufacturing high-quality goods.

3. Minimizing resource dependence

Beyond direct emissions, many businesses rely on the addition of materials and resources that require emissions to produce. By using software to determine how to use those resources more effectively, they are helping minimize emissions throughout the supply chain.

In agriculture, digital monitoring and planning can more efficiently deploy fertilizers while increasing crop yields. Similarly, steelmakers are reducing dependence on mined alloys by more than a third by using new machine learning technology. These efficiency improvements make existing resources go farther, helping lessen the environmental impact of production.

Software has prompted efficiency improvements in many sectors. For example, Amazon uses AI-powered software to squeeze efficiency from every corner of its logistics system, with inventory management algorithms that are continuously upgraded based on real-time data. The company’s warehouses are a model of efficiency — leading to higher profits.

Applying the Amazon model to sectors like manufacturing, shipping and energy, we could see companies using software to make their processes more efficient, eliminating waste and optimizing for energy use and resources. With increased efficiency translating into increased profitability, software adoption would benefit a company’s bottom line, as well as environmental impact.

Across industries, we should see software as an immediately applicable solution to be used with longer-term efforts to combat climate change. As carbon capture and similar technologies under development face the obstacle of costliness, software has the opposite effect — eliminating waste, optimizing energy consumption, minimizing resource dependence and overall driving efficiency improvements that translate into higher profits. With more money in their pockets, executives may be more likely to sign onto future green initiatives.

Berk Birand is cofounder and CEO at Fero Labs