The global sensor market is worth $230bn pa and likely accelerates due to the rise of AI. This data file has compiled 15-20 examples of AI systems integrating sensor data, as patented by industrial companies in 2024, to estimate what types of sensors, will be used in which contexts, and whether AI demand will surprise to the upside?
The motivation for the work was noticing that TE Connectivity, the manufacturer of sensors, connectivity and fiber optic systems, had unexpectedly featured in several recent TSE research notes, into say gas pipeline integrity or solar trackers.
Indeed, when we reviewed TE Connectivity’s own AI-related patent filings (in the data-file), we found it was integrating pressure sensors into the grips industrial robots, integrating imagery to automate manufacturing and warehousing, and using sensors for predictive maintenance, or to optimize HVAC systems.
Another HVAC specialist, BrainBox AI, which came up in our patent screen, was acquired by Trane in January-2025, with a technology that can reduce HVAC energy use by 25% and GHG emissions by up to 40%, using AI to predict and optimize the HVAC needs across different zones within commercial building spaces. HVAC is 40% of a commercial building’s total energy use.
There was good support for the idea that AI applications will pull on the demand for sensors: from JLL optimizing waste collection schedules based on sensor data from bins; or Halliburton and Aramco improve safety at well sites; helping Epiroc to electrify mining equipment; helping Krones affix labels to 90k plastic bottles per hour; and helping Raytheon and Airbus to attribute hard-to-diagnose equipment issues within aircrafts.
These case studies illustrate how demand for AI compute is likely to surprise to the upside. Most of us going about our everyday lives have no idea about the very specific challenges faced by industrial companies, where a solution can be extremely valuable, but requires smartening up a historically “dumb” processes.
Consumer products, of course, may also be improved by incorporating AI and sensor data. Sensors and AI feature in patents from Illy seeking the perfect espresso grind, from Gillette to improve shaving, and from Mars to predict if your dog is getting fat.
Visual images came out as the most common input data source, which is telling, as we were mainly looking for systems fed by pressure, temperature, motion sensors etc. Humans rely first and foremost upon vision. So maybe AIs will too.
Examples are given in the data-file, and make for a nice series of case studies, into how AI will increase demand for sensors. For VC/PE decision-makers, some interesting early-stage companies also came up in our patent screen, from improved EV charging to smart agriculture.