EDACafe Industry Predictions for 2024 – Oculi

By Charbel Rizk, PhD., Founder and CEO, Oculi

Charbel Rizk

Technological Innovations:

1. What major technological advancements or innovations do you foresee occurring in your industry in 2024?

In 2024 we foresee an increased demand for visual intelligence at the edge. The term smart sensor has been overused in previous years to mean adding or extracting some intelligence after the image has been captured, stored, or transmitted off the image sensor. While the result may indeed derive some intelligence, it is a highly inefficient process which is slow, power hungry, and expensive if the industry continues with the traditional dumb image sensor. Increasingly we see this demand for smart sensors not just as wishful thinking, but governments and regulators as turning it into policy requirements for certain systems. For example, in Europe, under the General Vehicle Safety Regulation, all new types of vehicles must be fitted with an advanced driver distraction warning system from mid-2024, where driver eye movements are monitored and warned when distracted to help reduce accidents. No doubt this will soon follow in the US and elsewhere. A new type of sensor capable of visual intelligence will be necessary to meet this and similar smart sensors or devices based on machine vision.

2. How do you plan to adapt or lead in these areas?

Oculi has had a head start and innovated on our technology over decades since we understood the problems with the conventional image sensor model of ever-increasing data from dumb pixels, high bandwidth transfer to power hungry post-processing processors to achieve current smart sensors.

Oculi is revolutionizing computer and machine vision for the AI age by transforming the image sensor from being dumb and rigid to being smart (i.e. vision) and programmable starting at the pixel, it is now capable of outputting visual intelligence (i.e. relevant information) instead of just data. Our novel computer/machine vision architecture provides a fast, accurate and reliable solution that outperforms the competition. Oculi’s architecture integrating neuromorphic sensing and processing at pixel level is ideal for edge applications (i.e. low SWaP & cost). The core innovation of our product, the Oculi SPU (Sensing & Processing Unit), is a smart programmable vision sensor that shifts the pixel from imaging to vision and enables fast, efficient, and secure 2D & 3D vision at any wavelength. It is capable of dynamic configuration to output select data in various modes, and with real-time programmability of output selectivity, sensitivity, dynamic range, and bit depth. By enabling continuous optimization, computer/machine vision deploying the Oculi SPU, in lieu of image sensors, can reduce the latency-energy factor by more than 600x at a fraction of the cost. Oculi SPU is sensor agnostic and can work with conventional CMOS, DVS, infrared, depth/TOF, and multimode.

In fact, Oculi has solved the decades-long data deluge challenge by developing a novel vision sensor that applies selectivity to process only the most relevant information or actionable data resulting in > 30X faster, < 1/10th the power drain, and protects privacy / data security at the sensor, all for a fraction of the cost. The net benefit of Oculi SPU is a reduced latency-energy factor of < 600x at a fraction of the cost of related products, and with unique privacy protections. In addition, as a Software-Defined Vision Sensor, the Oculi SPU cost is fixed, but the pricing model is incremental based on sensor speed and functionality, allowing customers to deploy the same hardware product for applications in multiple markets.

Market Trends:

1. How do you predict consumer behavior and market trends will evolve in 2024?

In 2024 and beyond, consumer and application use cases involving vision or detection devices will continue to demand more smart solutions in things they interact with every day. Despite all the intelligence we’ve heard about AI or your favorite smartphone assistant, those devices are not able to answer seemingly simple questions for you and I, which remain challenging for computer/machine vision. The automated faucet that won’t detect your hand to dispense water or the automated sliding door that keeps opening when you have no intention of going through it. Consumers are also looking for more smarts in mobile, gaming and AR/VR devices with visual intelligence that enables looks and feels more natural without draining batteries too fast.

2. What strategies are you implementing to meet these changing demands?

Oculi is shifting the edge from imaging to vision at the pixel. Through our innovation in the Oculi SPU, every pixel is a smart pixel making our product a smart programmable Vision Sensor, which results in a low bandwidth solution that is fast, energy efficient, and secure at a fraction of the cost. This is a paradigm shift from image sensors to the only Vision Sensor that can meet the always-on market requirements for all metrics (noted earlier) to scale the application use cases mentioned above that are not possible today. We are working with customers and partners to integrate the Oculi SPU into specific application use cases for face detection, eye tracking, gesture tracking, and others

Artificial Intelligence Integration:

1. How do you envision the role and impact of artificial intelligence in your industry in 2024?

Just as 2023 was the year that the world discovered AI through tools such as ChatGPT from LLM for human-like conversations, 2024 will be the year that the AI transformation continues into other deep tech/hardware industries including imaging. AI will become an important part of computer/machine vision as today’s dumb image sensors attempt to derive intelligence out of the tons of images that are captured. The problems that exist today of the deluge of data that must go through post-processing will not go away even if conventional image sensor manufacturers stick a label on those sensors as being “AI-enabled.” Certainly, AI will have an impact, but it will be diminished if the underlying problems with data deluge, latency and inefficient vision are not addressed.

2. What steps is your company taking to integrate AI into its operations or offerings?

Our company was built from the ground up to address the fundamental problems of today’s conventional dumb image sensors. Our innovative architecture integrating neuromorphic sensing and processing at pixel level in the Oculi SPU is in fact how we put the Eye in AI to enable visual intelligence and bring computer/machine vision a step closer to the efficiency of human vision. If you think about it, most human civilization and intelligence is built on our ability to process massive information visually. Despite tremendous, even astronomical advancements in technology (i.e. Moore’s law), human vision today remains 40,000 times more efficient than any machine vision. Our ultimate vision is to bring Visual Intelligence to many more devices with Oculi SPU to enable a new universe of possibilities in applications and use cases, thereby putting the Eye in AI!

Dr. Charbel Rizk is the Founder and CEO of Oculi, a fabless semiconductor startup with a mission to put the “Eye” in AI.  Charbel is a thought leader, Systems Engineer, Principal Investigator / S&T manager with expertise in drones, machine learning, multi-modal semiconductor sensors, sensor fusion, and perception for autonomy. Before founding Oculi, he also made key contributions at Rockwell Aerospace, McDonnell Douglas, Boeing, APL and Johns Hopkins University.

Category: Predictions

This entry was posted on Friday, January 19th, 2024 at 5:24 am. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.