The consequences of vulnerabilities within autonomous navigation are alarming

Autonomous navigation is entering new environments where the stakes have never been higher

Jennifer Edis Brand Marketing Manager Racelogic

When it comes to the trends and challenges facing autonomous navigation, there is one distinct factor that unites its application across transportation, robots, agriculture, space, marine, UAVs and drones, and that is an unprecedented rate of change.

The ancient Greek philosopher, Heraclitus, told us that ‘change is the only constant in life’, while renowned theoretical physicist, Stephen Hawking, told us that ‘intelligence is the ability to adapt to change’. So, if we are to be guided in the future by the wise minds of the past, we must accept that change is going to continue to drive the trends in autonomous navigation, and that intelligent thinking is going to be required to overcome the challenges.

New environments and demand for accuracy: The impeded performance of autonomous navigation within GNSS denied environments is no longer an acceptable limitation for modern technology. The ability to efficiently deliver accurate and reliable autonomous navigation in all conditions will need to become the industry norm. From advances in mass-market urban autonomous vehicles to biomedical autonomous robots within the human body, autonomous navigation is entering new environments where the stakes have never been higher.

Investment and collaboration: It is almost impossible to name an industry or company that does not include economic pressures within the list of challenges it is currently facing. When operating in a technological sector where keeping pace with advancements is not negotiable, finding creative solutions to the economic squeeze on an R&D budget is essential. From an accountancy perspective, the choice between inhouse development or collaboration is relatively simple. The opportunity to minimise risk in expenditure through collaboration is a highly attractive solution.

However, collaboration is not without its own challenges and organisations that are fierce competitors are now finding themselves in partnerships in a bid to stay ahead of, ironically, the competition. The technical partnership between BMW and Daimler is a prime example of this within the automotive industry which has seen a boom in collaborations in recent years as brands race to bring commercially viable autonomous vehicles to the road.

Several organisations are hedging their bets by combining in-house development and collaboration including some that are native to sectors far removed from autonomous navigation. Online retailer Amazon has directed significant spend towards finding viable solutions to the problems posed by last-mile delivery which has led them to autonomous delivery robots whilst additionally embarking in an autonomous vehicle collaboration with Toyota and investing in autonomous tech developer Aurora Innovation.

As autonomous navigation continues to mature, the technology options will narrow, and collaboration potentially followed by consolidation is likely to become the norm.

Regulation, legislation and ethics: The rate of change referred to in relation to autonomous navigation is reflective of the fast pace being set by technological advances and applications. However, when rate of change pertains to regulatory and legislative change, the pace can usually be described as pedestrian at best. Add to this the complexities of driving coordinated change at both micro and macro levels and there is clearly trouble ahead as this disparity in pace increases.

The ethics of autonomous navigation are still very much a work in progress when it comes to applying technologies to the real world, where human life depends on the decisions being made by the AI driving the navigation. The question remains, are users as well as bystanders going to be willing to accept the outcome of actions dictated by a software code?

It is widely accepted that self-driving technology has the potential to dramatically reduce death and injuries on roads but defining ‘how safe is safe enough?’ during testing and commercial rollout remains vague. Currently there is disparity between the many companies that are developing autonomous vehicles with differences showing between the approaches of traditional automotive manufacturers compared to new entrants to the sector. Uber recently unveiled its Safety Case Framework which outlines how the company has updated its approach to safety following a fatal crash in March 2018 involving one of its self-driving vehicles. Updating what an individual company defines is acceptably safe throughout the live testing programme contrasts with the approach of many traditional car brands who are looking to keep their autonomous profile blemish free following turbulent years of safety recalls and environmental issues.

The role of ethics surrounding autonomy has the potential to drive substantial changes to society and the merger of ethics and engineering is set to continue throughout 2020.

Security, jamming and spoofing: GNSS jamming and spoofing has been a hot topic throughout 2019, from Pokémon Go enthusiasts using cheap SDRs (Software Defined Radios) to catch rare Pokémon, to the US Department of Transportation’s Maritime Administration suspecting Iran of operating GPS jammers in the Persian Gulf to intentionally cause ships and aircraft to stray in to Iranian waters or airspace.

The consequences of vulnerabilities within autonomous navigation are alarming and it is understandable that the development of constellation authentication systems, such as Galileo’s OS-NMA (Open Service Navigation Message Authentication), and research focused on commercial applications of using dual-polarised antennas are being well received.

As with IT security, combating jamming and spoofing is a constant game of cat and mouse between those developing security systems and those intent on defeating them. However, the reality is far from a game (even when spoofing Pokémon!) and 2020 will need to ensure that GNSS technology remains an incontrovertible service to autonomous navigation.

Human factors: 2020 will be the year that human evolution will need to align itself with the autonomous revolution and the importance of the human factor should not be underestimated. Autonomous technology is not being designed or operated in a vacuum of humankind. Therefore, any system can only be as successful as its inherent human error and bias (unconscious or otherwise), and its aptitude for seamless human interaction.

The human factor also encompasses the changes to skills, education and knowledge required to both continue the advanced development of technologies and for society to prepare itself for the rapidly changing employment landscape. The most in-demand specialties in the job market today did not exist ten or even five years ago and the pace of change is set to accelerate.

Testing solutions: The common thread that ties together all other trends and challenges facing autonomous navigation is the need for testing solutions. New environments, new and collaborative technologies, changing regulations, security threats and human factors will all require testing solutions that can deliver irrefutable results without causing delays to competitive delivery schedules.

Testing autonomous vehicles has both degrees of difficulty and cost which must be weighed up against the returns required to maintain a feasible business model. As companies move towards testing Level 3 and above autonomous technologies, the challenges posed by testing and validation become considerable.

Using a test track to test Level 3 ‘eyes off’ vehicles is achievable with current technology, but the budget required to set up the physical testing parameters, run the programme for sufficient time and mileage, and then repeat for permutations in weather and driving conditions is vast.

Testing complete autonomy using controlled and geo-fenced test areas with no human presence already provides an achievable and affordable solution but having the technology and costeffective solutions to migrate this to ‘autonomy anywhere’ scenarios is not likely to happen in the near future.

Smart investment in testing solutions that are as complex and specialised as the technologies they are charged with scrutinising, is something that should be a priority for all in 2020. Many companies are betting their future on developing and testing autonomous vehicles and although it is unlikely that we will see the outcome of the wager in 2020, we might witness a change to the odds of success.