Next-generation navigation technologies for autonomous vehicles
One of the concepts that will drive the paradigm change in mobility is the Connected Autonomous Vehicle (CAV). Massive investments on the field and the latest advancements in Artificial Intelligence (AI) and sensors has moved relevant market uptakes for autonomous driving from 2035 to 2020.
CAVs are equipped with a huge number of sensors that allow them to understand the environment and act accordingly. However, this technology is superfluous without knowing the location of the vehicle in real time. Technology used to position a mobile device on earth is known as Global Navigation Satellite System – GNSS (e.g. GPS or GALILEO). Despite it seems impossible, currently, there are not any GNSS solution that meet the requirements of vehicle manufacturers for autonomous driving, due to:
- excessive cost to be implemented at scale (low margin sector)
- unavailability to provide location updates in real time under hostile GNSS conditions (e.g. urban canyons) and
- lack of a reliability measure to detect when a location is not accurate enough.
At Albora, we have built and patented the Albora Correlation Engine, which uses AI and, in particular, biologically inspired Deep Learning Networks to achieve the performance required by the sector. Moreover, our technology can be embedded on the electronics currently available on autonomous vehicles, allowing us to keep the costs extremely low (no additional HW required!)
To exploit our product, we plan to build SW packages of our algorithms and sell licenses through an easy to use API (SW company approach). This model is highly scalable and will allow us to tackle the huge market opportunity. In fact, SW will keep the largest market share for CAV, growing from €0.5 billion at 2015 to €25 billion in 2030. To this end, we need to assess the technical risks of migrating our code to more efficient programming languages, seek industrial partners to perform large pilots and fine-tune our business model using design thinking techniques.
Improved geolocation for autonomous vehicles
As technology advances, vehicles are becoming increasingly automated to improve safety and driving efficiency. An EU-funded initiative has developed the next generation of navigation technologies that are crucial for connected and autonomous vehicles (CAVs).
There is no doubt that the emergence of CAVs will revolutionise current patterns and modes of transportation. Major investments in the field and the latest advances in artificial intelligence (AI) and sensors means that significant uptake of autonomous driving is expected by 2020, rather than 2035 as previously forecast.
CAVs are fitted with an enormous number of sensors, enabling them to understand the driving environment and act accordingly. However, all this technology is meaningless if the location of the vehicle in real time is not known. Thus, global navigation satellite systems (GNSS), like GPS or GALILEO, are used to position a mobile device.
To date there is no GNSS solution that meets the stringent requirements of CAV manufacturers. This is partly due to the excessive costs incurred for implementing at scale and partly the inability to provide location updates in real time under difficult GNSS conditions, such as with urban canyons. An additional challenge is the lack of a reliability measure for detecting location accuracy levels.
Algorithms provide a software solution
The Horizon 2020 ALBORA project addressed these challenges, building and patenting the Albora Correlation Engine (ALCORE). The engine employs AI and biologically-inspired deep learning networks to achieve the required level of performance.
The initiative provided high-precision and ultra-low power geolocation technology and solutions for CAVs. “Our technology can be embedded in the electronics currently available on autonomous vehicles, allowing us to keep the costs extremely low, since no additional hardware is required,” says project coordinator Anselm Adams.
Researchers developed a software solution for the ALCORE. “This proprietary algorithm combines biologically-inspired architecture with advanced engineering and mathematical techniques that provides our clients with an accurate, fast, safe and reliable location coordinates,” Adams explains.
A huge opportunity
To better exploit their product, ALBORA will sell licences through a simple to use application programming interface – a software company approach. According to Adams: “This model is highly scalable and will allow us to tackle a huge market opportunity.”
In fact, software will comprise the largest market share for CAV, growing from EUR 0.5 billion in 2015 to EUR 25 billion by 2030. “To this end, we need to assess the technical risks of migrating our code to more efficient programming languages, seek industrial partners to perform large pilots and fine-tune our business model using design thinking techniques,” observes Adams.
Moreover, ALBORA provides a better understanding of the business side of the CAV sector through a comprehensive analysis of the market. “We believe all car manufacturers can benefit from our technology, which can provide centimetre precision for their CAVs,” Adams concludes.