Summary of "CES 2026 Keynote with Caterpillar's Joe Creed"
Overview
Caterpillar framed itself at CES 2026 as the “invisible layer” of the modern tech stack — the physical infrastructure (construction, mining, power) that underpins digital services — and described how it is embedding connectivity, AI, autonomy and analytics into that layer.
- CEO Joe Creed delivered the keynote emphasizing how physical infrastructure (roads, mines, data center power) is becoming data-driven and responsive via sensors, connectivity, software and AI.
- The company presented an integrated strategy combining edge AI, digital twins, autonomous machines and a cloud-native data platform to improve safety, uptime, productivity and sustainability.
- A safety-first approach and rigorous reliability/trust controls were repeatedly emphasized.
Key technological concepts and platforms
- Invisible layer: physical infrastructure instrumented with sensors, connectivity and software to become responsive and data-driven.
- Edge AI / on-machine intelligence: perception, planning, control and speech recognition executed on the machine to provide low latency and offline capability.
- Digital twins & AI factories: fleet sensor data is ingested to retrain models and redeploy improvements across fleets.
- Multimodal AI agents: coordinated agents handling speech, text, images and video to perceive, reason, plan and act.
- Safety-first design: pairing rapid innovation with stringent reliability, safety and trust controls.
Products, features and demos
Helios
- Cloud-native, event-driven digital platform unifying data from ~1.5M connected assets.
- Capabilities: handles thousands of messages/sec, millions of data pipelines/day, and stores >16 PB of reusable, high-quality data.
- Serves as the foundation for AI, analytics and services.
CAT AI assistant (“Cat”)
- Multimodal assistant built on Helios and edge compute; presented as a single interface (voice/chat) for operators and technicians.
- Demonstrated capabilities:
- Voice/chat queries for fuel efficiency, maintenance schedules, parts lists, utilization and warning light diagnostics.
- Proactive notifications and recommendations (for example, topping up generator fuel before storms).
- Field guidance for technicians (walkthroughs, parts provisioning) and in-cab co‑pilot tips for operators.
- Safety demo: operator sets an overhead “ceiling” (e.g., 13 ft); the machine automatically slows/prevents movement near power lines.
- Stated to be going live this quarter with in-field application validation underway.
Thor (on-machine platform)
- NVIDIA-powered on-machine platform that runs advanced models and control logic for robust edge AI in remote or off-grid locations.
- Designed to support energy-efficient models for imperfect connectivity environments.
Autonomous machine portfolio (construction)
- Previewed autonomous versions of wheel loader, dozer, haul truck, excavator and compactor.
- Intended to orchestrate jobsite workflows and improve safety, consistency and productivity.
Autonomous mining
- Longstanding, proven Level‑4 autonomy deployed for decades in mining.
- Fleet stats cited: >11 billion tons moved and >385 million km autonomously with no reported injuries.
- Customer case: Luxstone quarry driverless fleet hauled >2M tons.
Partnerships & enabling technology
- NVIDIA: collaboration on robotics, edge AI and accelerated computing (NVIDIA Isaac, Jetson, Thor). Focus on running perception/planning/control at the edge and updating models via a central AI factory.
- Customer collaborations showcased, including Luxstone quarry and an upcoming WM landfill demo at the Caterpillar exhibit.
Deployment, timelines & operational notes
- CAT AI assistant demoed live and said to be “going live this quarter.”
- Emphasis on edge-first deployment: on‑machine models and energy-efficient inference for areas with limited connectivity.
- Strategy centers on integrating hardware and software to improve safety, uptime, productivity and sustainability.
Workforce and social implications
- Caterpillar announced a $25M commitment for workforce training, education and partnerships to reskill workers for AI/autonomy roles (operators → fleet/site orchestrators, technicians, data scientists).
- Repeated emphasis that people remain central and that machines are intended to augment roles rather than simply replace workers.
Tutorials and usage guidance highlighted
- Operator/technician voice command examples (how to ask “Hey Cat” for fuel efficiency, maintenance parts, symptom diagnosis).
- Walkthrough of setting and enforcing a geofence/ceiling for safety in the cab.
- Technician guidance: parts lists, repair steps, and scheduling service with a local dealer.
Speakers and sources
- Gary Shapiro — Executive Chair and CEO, Consumer Technology Association (intro)
- Joe Creed — CEO, Caterpillar (keynote)
- Auggie (Auggie) Redzig — Chief Digital Officer, Caterpillar (Helios & CAT AI assistant demo)
- Jamie Miner — Chief Technology Officer, Caterpillar (autonomy preview)
- Deepu (Deep Tala) — Vice President, Robotics & Edge AI, NVIDIA (partnership remarks)
- Mark Perks — Caterpillar booth demonstrator (Cat 306 mini excavator demo)
Category
Technology
Share this summary
Is the summary off?
If you think the summary is inaccurate, you can reprocess it with the latest model.
Preparing reprocess...