This wasn’t an average CES, netting just over 100,000 attendees compared to almost a quarter million before the pandemic. However, it was much more normal than last year when attendance peaked somewhere in the low five digits.
But as for the reason technologists attend CES – to see bleeding-edge tech from name brands and disruptive startups firsthand – this year’s show was on-target as ever. Key trends on display included the move towards software-defined vehicles, the Matter standard’s coming out party, a maturing metaverse, and a shift in the UI development landscape.
You can experience them as if you attended yourself in the Floored: CES 2023 slideshow below.
1. And the Best Demo of CES 2023 goes to … TDK VibeSense 360
Embedded Computing Design’s “Best of CES 2023” demos went to TDK Corporation, whose VibeSense 360 IMU/motion sensor and machine learning software stack was showcased in TWS applications such as:
Spatial audio, shown by moving the “audio source” around a user in a mobile app
Transparency for Talk, which detects user voice vibrations to cut audio feeds so they can communicate clearly with those around them
2. And the Best Booth of CES 2023 goes to … Dassault Systèmes for the human body metaverse
While the rest of the industry is talking about digital twins and the future metaverse, Dassault Systèmes has already built it. They did it with the human body, where they are modeling human organs and simulating surgeries on them to help surgeons at Boston Children’s Hospital treat patients in need of life-saving care.
3. Let the Software-Defined Vehicle Revolution Commence
The biggest trend from CES 2023 that’s a matter of when, not if, is the transition to software-defined vehicles, shown in this graphic from Real-Time Innovations.
If you need proof, see recent acquisitions of Elektrobit by Continental, Wind River by Aptiv, ZettaScale by TTTech, and others. Is there an imminent shakeup of the automotive value chain on the horizon?
4. Starting the ignition for BlackBerry IVY
Years in the making, BlackBerry IVY now runs inside vehicles like this Jeep Grand Cherokee to execute predictive maintenance algorithms against data from brake pad and tire pressure sensors.
It also works with Amazon’s Alexa, enabling hands-free driver interaction.
5. Ambarella Expands CV3 Family of Automotive AI Domain Controllers
Ambarella, who specializes in Edge AI processors, recently announced the first production version of the CV3 family of automotive AI domain controllers, named the CV3-AD685. The part is aimed at Level 2+ to Level 4 autonomous vehicles.
Systems designed with the part were on display at the recent CES show, and I was privy to some of the demos, including a ride in a card equipped with some of the autonomous features. Being driven around with many of the previously human-controlled functions being operated by the car itself is still a little unsettling, at least for a grizzled veteran like me. But after a while, you (sort of) get used to it.
The CV3-AD685 works with the next generation of the CVflow AI engine, adding neural network processing that the company claims is 20 times faster than its previous generation CV2 SoCs. Additional enhancements include general vector processing capabilities that provide the overall performance required for full autonomous driving stack processing, including computer vision, HD radar, deep fusion and planning. It also integrates advanced image processing, a dense stereo and optical flow engine, Arm Cortex A78AE and R52 CPUs, an automotive GPU for visualizations, and a hardware security module (HSM).
The CV3-AD685’s neural vector processor (NVP) engine enables high performance, low latency, and low-power NN processing. In addition, the CVflow’s general vector processor (GVP) provides traditional computer vision processing while including specific optimization enhancements for HD radar. Twelve Arm Cortex A78AE CPUs and three dual-core, lockstep pairs of Cortex-R52 CPUs are included. CV3-AD685 is targeting ASIL-B on the chip level, with an ASIL-D safety island.
6. LIPS Demos Next-Gen 3D Stereo Camera
LIPS, a provider of 3D AI solutions, announced its LIPSedge S-Series Stereo Cameras at the recent CES show. And yours truly was on hand for a demo. I can say that it worked exactly as (or better than) advertised. Based on Amberella’s CV2 CVflow edge AI perception SoC, the next-generation 3D camera series fuses the power of high-resolution 2D and 3D image processing in one camera to enable applications that require the combined capabilities of 2D imaging and 3D machine vision.
The 3D stereo camera can support up to 720p depth and up to 4K RGB resolution and feature wide field of view, long range, and high accuracy. These RGB-D cameras are available in both USB and PoE versions and provide developers the high-quality results needed to build and deploy 2D and 3D machine vision applications.
Example applications include a robotic arm that can perform pick and place operations in 3D, along with visual inspection in both 2D and 3D. The camera can also enable service robots with the ability to navigate a warehouse in 3D while performing inventory control in 2D.
These LIPSedge S-Series features a global shutter sensor, which reduces visual distortion for superior perception in high-motion and vibration environments, and a 6-axis IMU, which is required for high-precision perception applications such as vision-guided robotics, autonomous mobile robots (AMRs), and unmanned drones. Samples of the new LIPSedge S-Series 3D Stereo Camera are available for evaluation now.