Periodic Reporting for period 2 - STV (Seeing the View)
Periodo di rendicontazione: 2020-03-01 al 2021-02-28
The automotive industry is amid a disruptive change highlighted by the entry of autonomous vehicles as well as electrification and ride hailing. While autonomous driving has made considerable strides in these past few years, current solutions are still not safe enough nor are they sufficiently reliable for commercial deployment, except in the most limited scenarios.
Most current OEMs and Tier 1 car manufacturing vendors are still looking for the “ultimate” autonomous driving technology that can identify any object as viewed by human eyes, while sustaining a reasonable cost for sensors and for the processing unit. The ability to detect the dynamic driving environment, such as vehicles, pedestrians and small obstacles on the road, while maintaining the affordability of the car is a daunting task.
VAYAVISION is a state-of-the-art sensor fusion software company. VAYAVISION’s leading environmental perception solution provides vehicles with crucial information on the dynamically changing environment around the vehicle in real time to enable safer and more reliable autonomous driving. The software solution encompasses state-of-the-art Raw Data Fusion with upsampling, AI and computer vision. While the common approach used by most platforms is ‘Object Fusion’, VAYAVISION’s ‘Raw Data Fusion’ brings an innovative concept to market that breaks the performance ceiling limiting the availability of self-driving cars.
VAYAVISION’s ‘Raw Data Fusion’ uses low-level data to construct an accurate RGBd 3D point cloud. Through its up-sampling algorithms, VAYAVISION increases the effective resolution of the sensors. This means that even low-cost sensors can be enhanced with VAYAVISION’s solution and provide high resolution understanding of the environment. VAYAVISION implements both Computer Vision algorithms as well as Deep Neural Networks in order to detect the various objects in the scene, including vehicles, pedestrians, bicycles, drivable road, obstacles, signs, and more. VAYAVISION also detects very small obstacles on the road with better detection rates and less false alarms than the legacy ‘Object Fusion’ solutions. Unclassified obstacles are also detected, providing an additional layer of safety to the vehicle.
In the STV project, VAYAVISION’s goal is to develop an architecture design, software design, and a demonstration unit as well as to sign 3 PoC projects with leading European OEMs and Tier 1 companies to evaluate the performance of VAYAVISION’s environmental perception software.
On the commercialization front, VAYAVISION’s goal is to study the market and correctly evaluate market pain points and requirements, as well as to understand the capabilities and product offerings of the competition.
During the STV project VAYAVISION has successfully finished designing the architecture and the software as well as implementing the solution in a real-time environment and building a fully operational demonstration unit. This demonstration unit was presented to multiple potential partners and customers (both OEMs and T1s) in Europe, Asia, and the US. These demos have resulted in 3 PoC projects as planned.
In the commercialization front, VAYAVISION has performed a market research and a competitive analysis in order to build a strategy for our technology and product. In addition, this research helped us better understand our competitive advantage and refine our value proposition.
Now, after finishing the STV project, VAYAVISION is ready with a state-of-the-art technology, a great demo kit that can show live demonstrations of our perception software to customers around the world, and a go-to-market strategy.
VAYAVISION has already started seeing the fruits of the work that was done as part of the STV project – we see the contract and commercial agreement that was signed with Trimble (a major US T1) to develop an automation kit for heavy machinery in the off-road use case, as a direct result of the work invested in the STV project. We are certain that the PoC projects that we have done together with the multiple demonstrations we have conducted, will results in more commercial agreements and contract with European and global OEMs and T1s during 2021.
Most current OEMs and Tier 1 car manufacturing vendors are still looking for the “ultimate” autonomous driving technology that can identify any object as viewed by human eyes, while sustaining a reasonable cost for sensors and for the processing unit. The ability to detect the dynamic driving environment, such as vehicles, pedestrians and small obstacles on the road, while maintaining the affordability of the car is a daunting task.
VAYAVISION is a state-of-the-art sensor fusion software company. VAYAVISION’s leading environmental perception solution provides vehicles with crucial information on the dynamically changing environment around the vehicle in real time to enable safer and more reliable autonomous driving. The software solution encompasses state-of-the-art Raw Data Fusion with upsampling, AI and computer vision. While the common approach used by most platforms is ‘Object Fusion’, VAYAVISION’s ‘Raw Data Fusion’ brings an innovative concept to market that breaks the performance ceiling limiting the availability of self-driving cars.
VAYAVISION’s ‘Raw Data Fusion’ uses low-level data to construct an accurate RGBd 3D point cloud. Through its up-sampling algorithms, VAYAVISION increases the effective resolution of the sensors. This means that even low-cost sensors can be enhanced with VAYAVISION’s solution and provide high resolution understanding of the environment. VAYAVISION implements both Computer Vision algorithms as well as Deep Neural Networks in order to detect the various objects in the scene, including vehicles, pedestrians, bicycles, drivable road, obstacles, signs, and more. VAYAVISION also detects very small obstacles on the road with better detection rates and less false alarms than the legacy ‘Object Fusion’ solutions. Unclassified obstacles are also detected, providing an additional layer of safety to the vehicle.
In the STV project, VAYAVISION’s goal is to develop an architecture design, software design, and a demonstration unit as well as to sign 3 PoC projects with leading European OEMs and Tier 1 companies to evaluate the performance of VAYAVISION’s environmental perception software.
On the commercialization front, VAYAVISION’s goal is to study the market and correctly evaluate market pain points and requirements, as well as to understand the capabilities and product offerings of the competition.
During the STV project VAYAVISION has successfully finished designing the architecture and the software as well as implementing the solution in a real-time environment and building a fully operational demonstration unit. This demonstration unit was presented to multiple potential partners and customers (both OEMs and T1s) in Europe, Asia, and the US. These demos have resulted in 3 PoC projects as planned.
In the commercialization front, VAYAVISION has performed a market research and a competitive analysis in order to build a strategy for our technology and product. In addition, this research helped us better understand our competitive advantage and refine our value proposition.
Now, after finishing the STV project, VAYAVISION is ready with a state-of-the-art technology, a great demo kit that can show live demonstrations of our perception software to customers around the world, and a go-to-market strategy.
VAYAVISION has already started seeing the fruits of the work that was done as part of the STV project – we see the contract and commercial agreement that was signed with Trimble (a major US T1) to develop an automation kit for heavy machinery in the off-road use case, as a direct result of the work invested in the STV project. We are certain that the PoC projects that we have done together with the multiple demonstrations we have conducted, will results in more commercial agreements and contract with European and global OEMs and T1s during 2021.
In the 1st year of the project, VAYAVISION finished performing market research and analysis, designed and built a demo unit, and signed a PoC project with a leading German Tier 1. VAYAVISION also reached out to multiple OEMs and Tier 1 companies in Europe and around the world to better understand market needs, this activity resulted in submission of multiple SOWs for PoC projects that hopefully will materialize during year 2 of the project.
During the 2nd year of the project, VAYAVISION continued to demonstrate its demo unit to potential partners and customers and performed 3 PoC projects with 3 leading T1 companies.
As part of the activities during the 2nd year, VAYAVISION has also built a commercialization plan that is based on the information that was gathered during numerous conversations and demos to prospective customers as well as a market research and a competitive analysis.
The aforementioned efforts have resulted in several SoWs and RFQs for projects with major customers in Europe (and other regions as well). Couple examples of such business opportunities that are currently being evaluated are: (1) a perception system for a L4 automation solution for buses in airports with a major European OEM, a bus manufacturer and (2) a perception system for a L2+-L3 autonomous high-way pilot for private passenger vehicles with a major European T1.
During the 2nd year of the project, VAYAVISION continued to demonstrate its demo unit to potential partners and customers and performed 3 PoC projects with 3 leading T1 companies.
As part of the activities during the 2nd year, VAYAVISION has also built a commercialization plan that is based on the information that was gathered during numerous conversations and demos to prospective customers as well as a market research and a competitive analysis.
The aforementioned efforts have resulted in several SoWs and RFQs for projects with major customers in Europe (and other regions as well). Couple examples of such business opportunities that are currently being evaluated are: (1) a perception system for a L4 automation solution for buses in airports with a major European OEM, a bus manufacturer and (2) a perception system for a L2+-L3 autonomous high-way pilot for private passenger vehicles with a major European T1.
VayaVision's researchers and engineers have developed a raw-data-fusion architecture that achieves great perception results already in the demo stage while running in real-time on a very efficient HW. We compare our SW's detection and tracking capabilities using publicly available databases and benchmarks, on which we achieve great scores and show continuous improvement.
The potential impact of our environmental perception is very substantial: if our performance meets our expectations, it will allow for mass proliferation of both L2+ ADAS systems as well as L3 and above autonomous vehicles. With VayaVision's design for Functional Safety and our robust SW architecture, the autonomous features will increase the safety of both drivers, passengers, and Vulnerable Road Users (VRUs) and have the potential of saving countless lives.
As of now, based on our results on public benchmarks (#1 in fusion of camera and radar and a top 5 result in real-time fusion of camera, radar, and LiDAR on the Nuscenes database and benchmark) as well as based on the feedback we receive from prospective customers and partners after seeing our demo and/or after doing a PoC project with us – we can proudly say that our performance is state-of-the-art and is market ready and can run on real-time hardware platforms that are available today.
The potential impact of our environmental perception is very substantial: if our performance meets our expectations, it will allow for mass proliferation of both L2+ ADAS systems as well as L3 and above autonomous vehicles. With VayaVision's design for Functional Safety and our robust SW architecture, the autonomous features will increase the safety of both drivers, passengers, and Vulnerable Road Users (VRUs) and have the potential of saving countless lives.
As of now, based on our results on public benchmarks (#1 in fusion of camera and radar and a top 5 result in real-time fusion of camera, radar, and LiDAR on the Nuscenes database and benchmark) as well as based on the feedback we receive from prospective customers and partners after seeing our demo and/or after doing a PoC project with us – we can proudly say that our performance is state-of-the-art and is market ready and can run on real-time hardware platforms that are available today.