There are unprecedented opportunities at the intersection of AI, project management, and sustainability. Unfortunately, without proper execution, projects to implement AI come with complexities, costs and a carbon footprint that undermine sustainability goals. It is essential for project managers to learn to implement “green algorithms,” specialized AI constructs designed to both enhance operational efficiency and prioritize sustainability. A five-step framework of defining sustainability objectives, centralizing data, customizing algorithms, pilot testing, and full-scale implementation can guide project managers and leaders as they deploy green algorithms effectively

I

n 2020, Microsoft announced its “Planetary Computer” initiative, a monumental project aimed at collecting environmental data on a global scale and making it accessible for sustainability solutions. By partnering with environmental organizations and leveraging cutting-edge AI algorithms, Microsoft aims to tackle some of the planet’s most pressing environmental issues, from biodiversity loss to climate change. This forward-thinking project serves as more than just an example of corporate responsibility. It stands as a significant milestone, indicating a broader paradigm shift towards making AI and sustainability the backbone of modern project management.

In today’s project economy, where projects are the primary unit of work and value creation, traditional project management paradigms are undergoing radical transformation. Sustainability is no longer a “nice-to-have” but an essential criterion for success — both on the initial completion of projects, and as project deliverables continue to deliver benefits over their lifetime. As companies like Microsoft demonstrate, there’s an unprecedented opportunity at the intersection of AI, project management, and sustainability (in the social and financial realms as well as the environmental realm).

Unfortunately, without proper execution, integrating AI and project management to achieve sustainability goals can contribute to the very carbon footprint we aim to reduce. How can we simultaneously harness the computational powers of AI and meet sustainability objectives without tipping the scales unfavorably in terms of cost and environmental impact? This is where the “green algorithms,” specialized AI constructs designed to both enhance operational efficiency and prioritize sustainability, enter the arena.

Complexity, Cost, and Carbon
Project management has always required a delicate balance of scope, time, budget, and benefits. Measuring the environmental, social, and human impact of projects adds further intricacies. AI tools for project management have the potential to help address these concerns — but as AI implementation shifts from theory to reality, another layer of three key challenges has emerged: the complexity, cost, and carbon of deploying AI.

Complexity

Interdisciplinary demands. Melding AI and project management demands a group of experts from varied domains. Complexity is multiplied when you bring together machine learning specialists, data scientists, environmental scientists, and policy compliance officers, just to name a few.
Dynamic regulatory environment. Sustainability objectives aren’t static; they evolve under the influence of local, state, and international laws, cultures, and traditions.
Data overload. AI runs on data, and managing and interpreting multi-dimensional datasets ranging from energy use to waste production introduces even more complexity.

Cost

Direct costs of interdisciplinary requirements. The interdisciplinary human resources described above come with a price tag. Expertise in multiple areas adds up.
Lifecycle cost. AI models are living entities that need continuous nurturing. This includes periodic updates, fine-tuning to adapt to new data or conditions, and even complete overhauls to incorporate new standards. Achieving and maintaining sustainability certifications, such as ISO 14001 or LEED, involves not just initial assessment costs but periodic audits that add to the project’s financial burden.

Carbon

Carbon impact of interdisciplinary requirements. Travel for stakeholder meetings, transportation of hardware, and even the energy consumed in office spaces where project teams work must be accounted for.
AI-related carbon emissions. Machine learning algorithms demand considerable energy, elevating the project’s environmental impact through extensive computational power. The lifecycle of the hardware required for these complex algorithms — from manufacturing to disposal—also carries an environmental cost. Specialized hardware — GPUs and TPUs — while accelerating algorithmic processes, add to the project’s overall carbon footprint. A study by researchers at the University of Massachusetts Amherst, found that the emissions from training just one AI model can be as high as 626,000 pounds of CO2 equivalent — about five times the lifetime emissions of an average American car.

Enter the Green Algorithms
This is where green algorithms come into play. They integrate seamlessly with a range of AI methodologies, from machine learning that enables predictive maintenance to Natural Language Processing for stakeholder sentiment analysis, and even reinforcement learning for dynamic resource allocation. The end results are AI tools that are not only smart in their execution, but also sustainable in their impact on projects. Green algorithms help address the challenges of complexity, cost, and carbon:

Managing complexity. Modern project management tools and methodologies are increasingly geared towards easing complexity and facilitating integration. Many platforms, such as Notion and Forecast, now offer features that simplify the management of complex algorithms and data sets, turning them into actionable insights for better decision-making. Specialized training and consultation services are also available to help project managers acquire the necessary skills for implementing green algorithms effectively.
Managing cost. Implementing AI solutions requires initial financial price and specialized skills, which could be considered barriers to adoption. However, viewing these technologies as investments with long-term payoffs rather than a mere expense is essential. Energy savings from efficient data centers translate into significant reductions of operational expenses and lower compliance costs. Intangible gains such as reduced risk and increased brand value may also accompany the adoption of green algorithms.
Managing carbon. With green algorithms, you get proven operational efficiency that comes with automated solutions and realize significant savings through optimized resource usage, waste reduction, and compliance with sustainability regulations. Switching from an average data center to a more efficient one can reduce carbon footprint by as much as 30 percent and further improvements are on the horizon.

GE Research provides a case in point. In collaboration with GE Renewable Energy, they have developed a cutting-edge machine learning tool designed to optimize wind turbine logistics, potentially saving the industry billions of dollars over the next decade. The algorithmic solution focuses on optimizing various aspects of wind farm operations, from shipping and installation to ongoing maintenance. Cost, complexity, and carbon emissions were all managed or reduced.

With success stories like this one becoming common, the message for project managers worldwide is clear — it’s time to learn how to implement green algorithms.

A Five-Step Framework for Adopting Green Algorithms
A structured approach can guide project managers organizations in deploying green algorithms effectively. This five-step framework provides a starting point:

1. Define sustainability objectives.
Establishing well-defined sustainability objectives is a necessity. These objectives act as a roadmap, helping to guide the AI-driven solutions that the project will employ, whether you’re aiming for carbon neutrality, waste reduction, or social benefits. Leverage frameworks like the United Nations’ Sustainable Development Goals to identify sustainability objectives. Asana or similar project management software can be tailored to include sustainability metrics in your OKRs (objectives and key results).

2. Centralize data.
For green algorithms to function optimally in a project management context, a strong, centralized data foundation is vital. This ensures that the algorithms have real-time and comprehensive data to work with for making environmentally sound decisions. Use data management platforms or data lakes to store sustainability markers like energy usage, emissions, and waste generation. Open-source platforms like CKAN can be customized for sustainability data tracking.

3. Customize algorithms.
The essence of effective project management with green algorithms lies in customization. Projects often have unique sustainability challenges that generic algorithms can’t address effectively. Adapt pre-built algorithms to meet your sustainability objectives, whether it’s reducing emissions or enhancing energy efficiency. TensorFlow and Scikit-learn are machine learning libraries that offer pre-built algorithms that can be customized to meet your project’s unique sustainability criteria.

4. Pilot testing.
Before green algorithms can be integrated into the larger project management framework, their efficacy needs to be tested in real-world conditions. A well-executed pilot test provides insights into how well the algorithm serves the project’s sustainability goals and what fine-tuning may be needed. Use simulation techniques to model your project’s ecosystem, allowing you to rigorously test your algorithms under various scenarios. Simulation software like Simul8 can help you create a digital twin of your project, facilitating the pilot testing of your green algorithms.

5. Full-scale implementation.
After successful pilot tests, it’s time to integrate these vetted green algorithms into the full-scale project management process. This is where the algorithm moves from being a theoretical concept to a practical tool that helps the project meet its sustainability goals. Opt for a phased approach, gradually incorporating the algorithm into different aspects of the project while monitoring performance metrics closely. Software platforms like Jira offer functionality to track the implementation process across multiple departments or sub-projects.

Project managers are learning that integrating green algorithms into their projects catalyzes operational efficiency and fortifies their commitment to sustainability. At the same time, organizational leaders are seeing that the convergence of AI and eco-consciousness offers a roadmap for achieving business outcomes. With meticulous planning and a commitment to sustainability, green algorithms can help project managers and sponsors turn challenges into opportunities for meaningful impact. As projects increasingly become the backbone of organizational value, green algorithms offer a tangible and effective way to achieve both sustainability and AI.

 
Antonio Nieto-Rodriguez is the author of the Harvard Business Review Project Management Handbook, five other books, and the HBR article “The Project Economy Has Arrived.” His research and global impact on modern management have been recognized by Thinkers50. A pioneer and leading authority in teaching and advising executives the art and science of strategy implementation and modern project management, Antonio is a visiting professor in seven leading business schools and founder of Projects & Company and co-founder Strategy Implementation Institute and PMOtto. You can follow Antonio through his website, his LinkedIn newsletter Lead Projects Successfully, and his online course Project Management Reinvented for Non–Project Managers.
 
Ricardo Viana Vargas, Ph.D. is the founder and managing director of Macrosolutions, a consulting firm with international operations in energy, infrastructure, IT, oil, and finance. He has managed more than $20 billion in international projects in the past 25 years. Ricardo created and led the Brightline Initiative from 2016 to 2020 and was the director of project management and infrastructure at the United Nations, leading more than 1,000 humanitarian and development projects. He has written 16 books in the field and hosts the 5 Minutes Podcast, which has reached 12 million views.