AI Helps Deliver Better Concrete More Efficiently

concrete iot ai

The construction industry uses 13 billion cubic meters of ready-mix concrete every year. The logistics and quality assurance of this concrete depend on legacy technologies that are no longer sufficient for today’s needs. To address this, Caidio is introducing AI-based solutions to help the supply chain deliver better quality concrete more efficiently.

The Challenges of the Concrete Supply Chain

The ready-mix concrete supply chain process starts at the batch plant that produces the concrete (typically a wet mix in Finland). Transit mixers then deliver the concrete to the construction site. At the site, the concrete is poured or pumped into molds and left to cure and dry. What happens at each stage has an effect on the quality of the final product.

The ready-mix concrete process (Image: Caidio)

Many of the measurement methods used in concrete quality assurance date back to the 1970s or even earlier. Documentation is usually paper based, and phone calls remain the main communication channel. Outdated practices make it hard to control the quality of the product throughout the process.

This is becoming increasingly apparent as demonstrated by some recent worrying examples of concrete quality problems in Finland. For example, in Northern Finland, in the city of Kemijärvi, a newly built bridge had to be dismantled because of fragile concrete. In Turku, concrete strength deficiencies halted the construction of a large hospital extension; indeed, the problems were so serious that they added an extra year to the construction schedule.

Collaborating to Make the Industry Better

Aku Wilenius, CEO of Caidio

A group of construction industry forerunners decided to get together to attempt to solve the quality problems. In late 2017, the newly established DigiConcrete working group arranged a workshop to identify the challenges in concrete construction. The group believed that digitalization was the key to better quality.

The next spring, the group identified Artificial Intelligence (AI) and the Internet of Things (IoT) as potential solutions to the problems. Unfortunately, DigiConcrete could not find any companies offering AI solutions for the concrete value chain. That discovery led to the establishment of Caidio, a concrete intelligence startup.

“Pasi Karppinen, whom I already knew and whose firm was a member of the group, suggested that we should start up a company, one that would revolutionize the industry,” says Aku Wilenius, a former National Instruments engineer, who joined the team and became the CEO of that startup.

Experimenting with AI

Working with members of the DigiConcrete group, Caidio started an experimentation project to study and test the possibilities of using AI for concrete construction. The DigiConcrete project received funding from the Finnish government’s KIRA-digi program.

One of project partners was Congrid, a company offering a digital alternative to paper-based quality reporting. In the experiments, Caidio used Congrid’s mobile concrete log sheet to test how AI could be used to analyze it.

“Another test involved an AI-based construction assistant. As we know, knowledge in the industry is very much person-specific and we wondered how that knowledge could be passed on to a less-experienced worker,” Wilenius explains.

The idea was that the digital assistant could collect data across the entire concrete construction process, including from design models and databases, to open data sources, and IoT sensors. It could use weather and traffic data to determine the behavior of ready-mix concrete during transport. It was intended that AI algorithms would control the whole process and product quality, prompting users to make the right decisions throughout the process.

One of the experiments tested locational tags to track down concrete pumps on a construction site. Surprisingly, it’s not always straightforward for a truck driver to find the right pump on a large site.

The KIRA-digi experiment also included tests on concrete quantity estimation, mold heating during the winter, logistics optimization, raw material quality control, and safety on the construction site. In addition, Aalto University is developing new digital measurement methods and sensor technology for concrete quality measurement in collaboration with the DigiConcrete project.

“In 2019, we’re planning to pick the most feasible use cases from our experiments and to start piloting them in real-life construction projects,” says Wilenius. “As a company, we’ll focus on providing the underlying intelligence for all relevant processes. We’re also always happy to collaborate with others who want to develop apps and user interfaces that utilize our inference machine.”

You can connect with Aku Wilenius on LinkedIn.

Share This