Transforming into an analytics based organization We have been busy trying to democratize AI, data & advanced analytics within our organization including launching H.I.P. the Holcim Intelligence Platform that brings all business, innovation and tech stakeholders together in a collaborative online space. We believe engaging all employees—not just tech staff—in analytics development ensures that analytics solutions truly augment employees in their roles so they can do their jobs more efficiently, and it stimulates employee ingenuity, confidence, and flexibility.
Data and Algorithms versus Human Thoughts and Wisdom Will data and quantification replace human wisdom? I argue it will augment it through conscientious use and application. Data can help us make better decisions and free our time and energy to focus on the tasks where we can make the most difference, and, ultimately, make the world a much better place.
How is AI Powering Predictive Maintenance As the world's largest building materials company with 2000+ facilities (plants, factories, terminals, distribution centers, quarries etc.) operating in 70+ countries, we think about how to best reduce our overall maintenance cost at a local, regional and global scale and increase the lifespan of our equipment. Here I discuss how we are using AI to accomplish this at Holcim.
Scaling AI in Supply Chain As the world's largest building materials company operating in 70+ countries, we think about how to best optimize and manage our global supply chain using AI at Holcim. As an example of scale, we work with over 20,000 transport suppliers and 90,000+ drivers that travel approximately 1.7 billion kilometers for us each year moving our material. Here are a few lessons we learnt from our open innovation efforts scaling AI within our supply chain.
Using AI for Predicting order cancellations A good example of machine learning, data and business collaboration within my teams at Holcim. Algorithms to reduce waste, reduce thousands of driving hours and limit trucks on the street.
Predicting the Quality of Cement in Advance So that our homes, schools, bridges, roads, and tunnels do not crumble and collapse, the strength of the cement that builds them becomes a bit of a big deal. Cement must attain strict standards as far as strength development is concerned. Mortar prisms are being formed and stored under special conditions to perform the strength tests. Strength behavior is defined at Early Strength and as well as Late Strength. Late strength performance of a product is only known 28 days after it has been produced. With the aim to achieve the standards and norms required, cement producers like us are prone to add some safety margins on the product (either fineness demanding for more grinding energy and cost, or increasing the clinker factor and reducing the amount of inert materials (filler) such as limestone, using strength enhancers in the product, or a combination of all.
Using Machine Learning for reducing emissions Industrial production, manufacturing, logistics, and building materials are leading causes of Carbon emissions. Here I talk about how machine learning made an impact towards reducing emissions in the sector.
Involving business stakeholders using proactive outreach, inclusion and survey feedback Becoming a data driven organization requires contribution from our business and workforce. This is not something that data, digital and technology can do on our own in a vacuum, we need partnership with business to be able to do this. Data, hand-in-hand with focused change management plus well curated ideation sessions and team governance, help us identify smarter ways to conduct business, whether it is predicting equipment failure or forecasting customer demand.
Remote inspection of a factory using Virtual Reality This is a demo of a real time remote controlled VR system using data, VR and teleporting. My manufacturing analytics team built this VR system with our digital team that trains an operator to carry out a remote inspection of our plant (the example shown is that of a Kiln inspection). We have now deployed this across all our 2000+ plants worldwide.
Making sense of data using 10 different lenses or filters Data gets manipulated, mis-utilized, misquoted and misrepresented in ways that make us believe something that may not be factually correct. We need to understand the lenses through which we can make better sense of the data so we become smarter consumers and citizens.