When it comes to autonomous vehicles, it is a no-brainer that one of the first questions that comes to mind is “How safe is this technology?”. It pushes the boundaries of human innovation and prompts us to create new frameworks and approaches to how we think about safety. In this article we share a glimpse into FERNRIDE’s safety philosophy and outline why this is the crucial milestone on the path to industrialisation of autonomous trucking, starting with logistics yards.
FERNRIDE offers scalable automation solutions for yard trucking. Private yards in the logistics industry is where we see the biggest need, and the biggest potential for this kind of solution today.
The logistics industry grapples with slim profit margins, truck driver shortages, and the push to cut CO2 emissions. The truck driver shortage in Europe is set to rise from 400,000 to 2,000,000 by 2026. Rising labor costs and a shrinking workforce will put a further strain on profits. As the climate crisis intensifies, companies face stricter emission regulations and potential fines. As the logistics industry strives for transformation and more automation, it is clear that solutions must be swift, economically sound, and integrate seamlessly into current logistics operations without causing disruptions.
This is why at FERNRIDE our goal is to facilitate the transition from manual and diesel to autonomous, electric trucks. We focus on solutions for problems that can be provided today in confined areas such as logistics centers, production facilities, and intermodal and container terminals, starting to scale within the latter. Rather than solely relying on autonomous technology, we believe that human-machine collaboration will ensure economic viability and deliver the benefits our customers need today. We call this approach “human-assisted autonomy”.
In order to deliver the desired benefits today, along with the scalability and the highest reliability of operations, FERNRIDE employes the human-assisted autonomy, which allows for remote takeovers of trucks when necessary. Simply speaking, this way we can automate all routine routes and tasks, and ensure that there are no disruptions in unknown or complex traffic situations and maneuvering. This powerful combination of autonomous technology and human oversight create a solid reliability layer that only fully autonomous solutions cannot guarantee.
“Safety first” is not just a buzzword for FERNRIDE, it is one of our fundamental values in everything we do: from product development to our approach to operations. Creating a strong safety concept and implementing it, is also, naturally, a major step towards the industrialization and scalability of our solution.
Being a safety driver is an important job,a key safety fallback layer for pilot and R&D projects, and is a crucial component in the stages of proving the concept and operations adaptability. However, when it comes to the commercial ramp-up and scaling of the solution, a safety driver has to be taken out of the equation and replaced by a robust and sound safety case.
We build our Safety Case based on the high safety standards of the EU with regard to machinery and automation. It is our belief that this creates a basis which allows us to scale globally and efficiently adapt to complying with local product safety and product compliance frameworks.
What are the key elements of our approach to safety?
This way, we will reach CE conformity, which as such indicates that the product meets EU standards for safety, health, and environmental protection and allows the free movement of products within the European market.
First of all, a truly convincing safety case requires at its core a holistic safety culture within the whole team. This culture is what we are crafting as a prerequisite of everything else. Developing a robust and reliable safety case requires an experienced and knowledgeable team, a comprehensive process, and a strong collaboration between technical teams, engineering, and operations. And while FERNRIDE has, arguably, the strongest team in Europe with a proven track record of launching safety-critical products, let’s take a closer look at the process.
There are four main steps for designing a strong safety concept: definition of the operational design domain (ODD) and driving scenarios, hazard assessment and risk analysis, specification of the risk reduction measures, and, finally, re-estimation and re-evaluation of risks.
In a nutshell, this is a complete description of static and dynamic aspects of the driving environment. This includes, but is not limited to, the environmental conditions, cataloging of actors and objects, scenarios and required maneuvers, vehicle interactions with other actors in the yard, as well as vehicle properties. Our team creates a comprehensive description, taking into account every detail and special features of every use case and yard. For instance, containers lifted into the air by cranes or reach stackers are usually specific to container terminals and ports, and mixed traffic operations with people and manually driven vehicles on site add more complexity.
In the next step, a thorough collection and identification of all hazards and potentially dangerous situations starts. The guidance is provided by technical regulations and applicable standards, and at the core of it lies a general approach to risk assessment and reduction, specified in the ISO12100. Additionally, our teams think about all the possible external risks, as well as all the ways our system can malfunction, including the careful consideration of cybersecurity threats.
After that, every possible risk is documented and evaluated based on the severity, frequency of exposure, as well as probability of occurrence and avoidance. This is how an initial Risk Index is calculated.
There are three ways to address the identified risks: by changing the design of the machine, by building technical safety functions, or by creating operational measures. It’s important to point out that a technical measure is always preferred over an operational one.
Safety measures can address different aspects of risk: how severe it is, how often it happens, and how likely it is to occur or to be avoided. For instance, if we make a safety measure more rigorous, it's less likely to happen. In many scenarios slowing down can make it easier to avoid accidents. Some safety measures use sensors to detect objects, traffic participants, or people to avoid collisions and prevent accidents.
Creating a sound safety plan is crucial, but what comes next is the trickier step of the process. It has to work in all possible situations and even the unusual ones, while still being doable and acceptable in practice. This is where the testing, validation of safety functions, and the worst-case scenarios analysis come in. This process, when done thoroughly takes a lot of effort and is time-consuming: We have to understand the worst-case scenarios and test them in controlled conditions to make sure they work reliably.
After this work is done and the sufficient amount of testing has taken place, it is time for a re-estimation and re-evaluation of the risks. The simplified question to answer here: Are we complying with the required safety standards with all our risk reduction measures in place?
Performance and availability of the solution are two factors that should be looked at, along with the safety standards of the product. Uptime, reliability, scalability, and cost are crucial in the logistics industry, this is why a safe product, that causes constant disruptions of operations, won’t do so. To deliver the safest solution and the best customer value, it is important to aim for Key Performance Indicators (KPIs), like reliability, increase in productivity, return on investment (ROI), and, finally, scalability of the solution: Is the safety system available and functional in any weather conditions? Is the cost of the stack within the limits for a customer to scale the solution? What infrastructure changes are required, if any? How does the solution boost productivity and increase the efficiency of the operations?
FERNRIDE is committed to creating the new “gold-standard” system for automated trucking. This means that we develop the product according to the high standards based on compliance with applicable regulations (product compliance) as we’ll innovate on technology to achieve the best-in-class KPIs mentioned above that matter for the customer. We firmly believe that human-assisted autonomy is the only path to scalable and reliable automation of the logistics industry, particularly, in container terminals at ports. With this approach we deliver the desired benefits already today: increase productivity, safety, and create better jobs.
Tilmann Ochs, Director for Systems, Safety and Security at FERNRIDE
Tilmann Ochs received a degree in computer science from Technical University Munich in 2004. He then worked at BMW mostly on functional safety of compute platforms for highly automated and fully automated driving. From 2019 he worked as safety engineer at Autonomous Intelligent Driving GmbH, a startup company backed by Audi developing autonomous robotaxi systems. The company was later acquired by Argo AI LLC, an autonomy company backed by Volkswagen and Ford. In 2022 he joined FERNRIDE as Director for Systems, Safety and Security.
Dr. Volker Hartmann, General Counsel at FERNRIDE
Volker Hartmann has over 15 years of experience as an in-house legal counsel specializing in technology and product law in the automotive and mobility industries. He has advised autonomous and connected driving R&D projects, M&A projects, and built up legal teams at Audi, Daimler, the VW Group, as well as at several tech start-ups. As our general counsel, his duties will encompass among other matters legal & regulatory affairs, data protection, compliance and governmental affairs, as well as the establishment and expansion of our office of the general counsel.