Skip to main content
Helikon Labs
Sign inStart free

Who’s working on industrial AI and digital twins?

Plants, factories and fleets now stream more sensor data than anyone can read, and the work is in turning it into models that predict failures, catch defects and test a change on a digital twin before it reaches the line. This page shows how much of that work the Helikon Research Graph holds and gives you a problem statement to run in Partner Search.

Run this in Partner Search

What the Helikon Research Graph holds on industrial AI and digital twins

17,000+

researchers

46,000+

publications

1,000+

U.S. patents

Counts from the Helikon Research Graph on 13 October 2026, rounded down. Each figure is counted on its own, so they aren’t added together.

How these were counted
  • Publications: papers whose title, author names or research topics match the keyword set below, the same text the Literature search reads.
  • U.S. patents: granted U.S. patents whose text matches the keyword set and that carry at least one of these CPC subclasses: G05B, G06N, G06Q, G01M, G06F, G06T. That is the Patents search with the CPC subclass field set.
  • Researchers: people in the graph who authored a matching publication or are named inventors on a matching patent.
  • Keyword set: "digital twin" OR "digital twins" OR "predictive maintenance" OR "condition monitoring" OR "machine fault diagnosis" OR "fault diagnosis" OR "bearing fault" OR "remaining useful life" OR "prognostics and health management" OR "industrial AI" OR "industrial artificial intelligence" OR "industrial machine learning" OR "smart manufacturing" OR "Industry 4.0" OR "cyber-physical production system" OR "industrial internet of things" OR "IIoT" OR "soft sensor" OR "process monitoring" OR "statistical process monitoring" OR "anomaly detection in industrial" OR "industrial anomaly detection" OR "automated visual inspection" OR "surface defect detection" OR "machine learning in manufacturing" OR "AI in manufacturing"

Try it on a real problem

Partner Search ranks researchers against a problem statement by what their publications, patents and profiles say. This one’s ready to run. Edit it to fit your project.

We monitor a few hundred pumps and motors with vibration and current sensors, but our failure prediction raises too many false alarms to act on. Looking for a group with fault diagnosis for rotating machinery, remaining useful life estimation and models that transfer across machines with little labeled failure data.
Run this in Partner Search

Who’s working on this

Mechanical, industrial, electrical and computer engineering groups at universities, national laboratories, and company teams in industrial automation, energy, manufacturing, aerospace and enterprise software.

Universities and research labs

Ranked by researchers with work in this area, researchers whose main topic it is, and the U.S. patents they own.

  1. Georgia Institute of Technology

    298 researchers · 892 publications

  2. University of Connecticut

    24 researchers · 298 publications · 4 U.S. patents

  3. Carnegie Mellon University

    174 researchers · 214 publications

  4. Purdue University West Lafayette

    81 researchers · 286 publications · 3 U.S. patents

  5. Pennsylvania State University

    240 researchers · 435 publications

  6. North Carolina State University

    167 researchers · 284 publications

  7. University of Wisconsin–Madison

    149 researchers · 305 publications

  8. University of Michigan

    192 researchers · 396 publications

Companies

Ranked by the U.S. patents they own in this area and the researchers who work on it.

  1. General Electric

    45 researchers · 144 publications · 35 U.S. patents

  2. IBM

    59 researchers · 76 publications · 52 U.S. patents

  3. Ford Motor Company

    32 researchers · 82 publications · 8 U.S. patents

  4. General Motors

    40 researchers · 118 publications · 11 U.S. patents

  5. Boeing

    15 researchers · 43 publications · 11 U.S. patents

  6. Rockwell Automation

    20 researchers · 35 publications · 22 U.S. patents

  7. Honeywell

    19 researchers · 42 publications · 20 U.S. patents

  8. Johnson Controls

    6 researchers · 7 publications · 29 U.S. patents

From the Helikon Research Graph on 13 October 2026. Researchers are people now at the organization who authored a matching publication or are named inventors on a matching patent. Publications are matching publications that list the organization among their authors’ affiliations at the time of publication, counted once each. Patents are matching U.S. patents the organization owns. Each name opens the organization’s page on Helikon Labs.

Related

More technology areas

See all technology areas

Search the whole graph on the Free plan

Start free