Coming Soon: Oxa's 'From Here to AV Eternity' Report

• Oxa's CEO, Gavin Jackson, has worked in executive roles at global technology companies including Microsoft, Amazon Web Services, EMC, VMware, and UiPath. He graduated from Cranfield Business School, IMD School of Management, and Harvard Business School’s executive leadership program.
• In this article, Gavin introduces a major new Oxa report assessing how commercial autonomous vehicle (AV) deployments can reach massive scale, and the challenging questions all AV innovators must answer to ensure safe deployment of vehicles across all sectors and markets.

Scaling self-driving from a trickle to a torrent of vehicles

Many of our technical experts have been working on autonomous vehicles (AVs) for more than 25 years. Projects have included deploying self-navigating drones in freezing waters beneath the Arctic sea ice, and for Oxa - trialling road vehicles on countryside lanes in the UK and scorchingly hot test tracks in the US. All of this rich experience is part of the development journey the AV industry has been on.

In 2024, we’re facing off to a new deployment challenge; arguably the most hostile terrain any autonomous vehicle has ever encountered. The industry is entering the commercial world, where all AV companies are trying to prove their technology is robust from an engineering and a financial perspective.

Success for Oxa begins with the fundamental reasons why autonomy is needed on- and off-road. One of the main motivations is cutting congestion while enhancing societal connectivity and reducing emissions. A challenging trio to master. On the business operational side, companies are striving to extract greater value from transportation systems, mitigate severe driver shortages, and find the most cost-effective approach to deploying advanced technology so they can steal a march on competitors. All questions lead to the same answer: The smart adoption of autonomous vehicles (AVs) at scale.

A new passenger-carrying shuttle using Oxa Driver software being deployed by our US fleet operations partner Beep at Lake Nona, Florida.

There are now numerous companies working in autonomy as its value becomes more apparent. This year, Oxa expanded into international markets, and by the end of 2024, our self-driving technology will be in use in multiple passenger-carrying shuttles, an energy farm, global hub airport, and retail environment. We’ll have deployed Oxa Driver software in more than a dozen different vehicle types: AV shuttles, transit vans, 4x4 pick-ups, small industrial trucks and more. And we'll have seen Oxa powered vehicles having run in over 10 cities, across 2 continents; all on the same foundational software. It's breathless stuff but no company should ever get carried away!

Achieving commercially viable AVs in a production-line manner remains a work in progress for everyone. Disagreement remains over the sectors in which to deploy first, and there is no consensus yet on new laws creating a universally safe environment for AVs. However, momentum is picking up with new urgent needs for autonomy beginning to appear - mainly in passenger shuttling and industrial logistics operations. Our upcoming report, From Here to AV Eternity, publishing in the next few weeks, will lift the lid on those opportunities and assess how autonomy can reach maturity rapidly, safely and sustainably.

Differences of Opinion Abound

As I write, the initial group of a few dozen AV innovators is being narrowed down to those with the most effective business models and capabilities. This is normal for all new technologies as they mature. For instance, there is ongoing debate about the optimal AI approach. One of the most controversial aspects is the utilisation of mapping data. Maps in the AV domain are significantly more intricate than those used by non-autonomous vehicles, and many innovators have invested millions of dollars in their development.

The debate revolves around the level of detail required by an AV for a journey. One group argues that the vehicle only needs to know a journey's start and end points. The AI interprets the route and hazards in real-time using on-board sensors (with many relying purely on cameras). The challenge is that the AI, although highly sophisticated, learns in an unsupervised manner in real-time. For instance, it does not know that it should drive cautiously over traffic calming bumps until it encounters too many at high speed.

The second camp (Oxa included) believe in backing up AI interpreted sensor data with highly detailed maps and pre-defined driving rules governing what the vehicle is allowed to do. An AV map incorporates elements like street furniture, road architecture and elevation, even roadside trees that may obstruct a sensor's view. This forensically detailed information is supported by a sophisticated sensor suite, typically comprising Radar, Lidar, and HD cameras. Deciding on your approach depends on the level of trust you are willing to place in your AI system as it learns and evolves, as well as the cost you are willing to incur in preparation for your code hitting the road. And all of it also relates to how a company thinks about the importance of building trust.

Related to the trust issue, there is another debate about the extent of data to be shared with regulators in the event of an incident. Currently, data sharing primarily involves written incident reports and recorded information, such as videos, rather than the sharing of comprehensive system information. Essentially, the AV operator must describe the 'what' but not the 'why'. The challenge now is to devise a method of transforming the AV into a reliable expert witness, enabling the complete reconstruction of an incident at both a physical and detailed digital level to facilitate rapid learning. We believe that a more forensic approach to safety is feasible and essential, but it remains a subject of debate.

Let our New Report Be Your Guide to Autonomy

Follow Oxa on LinkedIn for updates on report publication

The AV evolution is undeniably complex and our soon-to-be launched report, From Here to AV Eternity, will lift the lid on all areas governing the commercial success of autonomy in transportation. It goes far beyond the technology and it can be your guide to the rollout roadmap with comprehensive commentary from leading autonomy sector experts.

Paul Newman, UK CAM Champion and Oxa Founder and CTO, discusses the various technology approaches. Silicon Valley company veteran and Oxa Chief Product Engineering Officer Graeme Smith answers questions about how cutting edge AI can be safely turned into off-the-shelf products. Additionally, less-explored areas are also covered, including insurance, investment, and strategies to facilitate job transformation. Reports being published already point to a wealth of new opportunities such as Steer Group's Fourth Economy assessment identifying the need for 190 new jobs in manufacturing and servicing for every 1,000 AVs produced. Other contributors include Jamie Vollbracht, Founding Partner at Kiko Ventures, Dr Nikolas Thomopoulos, Associate Professor of Transport, at the University of Surrey and others from Oxa with backgrounds ranging from technology to commerce to regulation and beyond.

The complete report will be released soon, click here to read the first instalment, authored by Oxa Chief Commercial Officer, Richard Jinks. Richard discusses how autonomous vehicles can move from R&D trials into profit-making deployments at scale. Amidst the excitement of technological advancements, it is easy to overlook the necessity for autonomy to become a viable business opportunity; otherwise, the numerous positive impacts will only benefit a small minority, and the full potential of AVs will remain unrealised. Part of the argument centres around the order in which deployments happen. You can look at my prediction below.

Oxa's view of how autonomous vehicle deployments will scale commercially

Follow us on LinkedIn for updates on the report’s publication, and all of our other news too.

Gavin Jackson
CEO, Oxa