By Josh Webb, Vice President of Business Development, Predii

AIThe rise of “Industry 4.0” is symbolized by newly connected devices and intelligent machines, and their presence has redefined what is possible. The IoT (Internet of Things), AI (artificial intelligence), and machine learning are offering the construction industry helpful predictive maintenance capabilities. These can help organizations to estimate forthcoming maintenance needs, and prevent unforeseen breakdowns and their resultant downtime. However, with connected sensors communicating equipment status to the cloud in real time, there are many further benefits to be gained.

By continuously recording a range of sensor readings across the organization, AI systems are able to learn what constitute the ‘normal’ operating baselines for different machinery. By monitoring for activity that falls outside of this range, these systems can generate alerts and actions to increase the performance of machinery, or mitigate potentially unsafe conditions that could go unnoticed by a human operator.

Consider a piece of heavy equipment, such as a truck or excavator, for example. As the engine develops wear, and various mechanical problems surface, its performance can be diminished resulting in lower fuel efficiency, less power output, and therefore slower progress on the job. Or, as its braking systems become worn, they can generate measurably greater heat, reduced stopping force, and other factors that can predict possible hazard situations (e.g. potential failure to stop as an increased load is taken on board). These issues can be pro-actively detected, and appropriate actions triggered to mitigate the risks with advance warning (maintain, repair, replace, upgrade). This will prevent the possible outcome of operating at reduced capacity due to breakdown, or increased physical danger due to mechanical failure.

Another scenario for this equipment is to monitor position for safety and causational factors. For example, if the equipment generates a rapid, unexpected start/stop movement, this could indicate that it has crashed, overturned, or some other anomalous movement in which a human operator may be in danger. On the other hand, if the equipment begins to routinely generate anomalous data like this, it could indicate that a human operator is performing their job in a dangerous fashion (rapid acceleration/braking, working on unsafe grades, etc.). This can generate a different kind of alert, such as alerting a manager to retrain such behavior, or define safer SOPs (standard operating procedures).

When safety, productivity, and total cost of ownership (TCO) are all priorities, it makes sense to use the best technology available to monitor the data generated by an organization’s assets. It is very possible to make practical use of IoT data across an entire enterprise today, and a specialized vendor can help to assess the available data and create a machine learning strategy for IoT equipment. With such an approach, the modern construction organization can see immediate benefits today, and future-proof itself for tomorrow.

This article previously appeared in the CONEXPO-CON/AGG 365 newsletter.