Edge Case

Edge Case is heading to Sea-Air-Space Conference this week. Across naval and defense systems, the same themes continue to surface: increasing system complexity, growing volumes of data, and the challenge of maintaining a clear, current understanding of risk. We’re looking forward to connecting with leaders across the community who are thinking about how safety needs to evolve alongside modern systems. If you’ll be there, let’s connect.

Most safety conversations around autonomous vehicles start with the same question... “Are they safer than human drivers?” Join Partners for Automated Vehicle Education (PAVE) on April 30 as our Chief Safety Officer, Michael Wagner, joins leaders from Waymo and TRC to explore how the industry is thinking about that question and where it starts to break down. The human driver baseline is more complex than it first appears. Safety is not a single benchmark or a lagging indicator. It is built from evidence across systems, including simulation, structured testing, and real-world operational data, and from how those signals are interpreted over time. We’ll explore topics around: - Why the human driver baseline is more nuanced than it sounds - How leading indicators can complement traditional crash-based metrics - What it takes to evaluate automated driving systems beyond direct human comparison - How safety evidence is communicated to build meaningful public trust As the industry scales, the conversation is shifting from “safer than human” to how we define and demonstrate “safe enough” in complex, real-world systems. Register to join the conversation - https://lnkd.in/ecaNBer8

Doubling system complexity does not double safety effort. It multiplies it. More interactions. More dependencies. More ways for assumptions to break. At a certain point, adding more people or more reviews does not solve the problem. Edge Case sees teams hit this wall as systems become more software-driven and AI-enabled.

Great SAE WCX experience. Appreciate everyone who participated in the four talks I was lucky enough to contribute to! Here’s a photo of me gesticulating to uplift safety technology :) Edge Case Carsten Thomas Michele Gittleman Paul Schmitt Chaitanya Shinde Jeffrey Wishart, Ph.D. #WCX2026

Safety evidence lives everywhere. Test results in one place. Operational data in another. Engineering changes somewhere else. When evidence is fragmented, confidence is too. Different teams make decisions based on partial views of risk. Edge Case works with organizations trying to unify how safety evidence is understood.

Safety is ultimately about understanding risk well enough to act. When that understanding lags behind the system, decisions are made with incomplete context. Sometimes that leads to delays. Sometimes it leads to overconfidence. Edge Case focuses on helping organizations close that gap between system reality and safety understanding.

We are thrilled to have our Chief Safety Officer, Michael Wagner, taking the stage today in Detroit at SAE WCX! As the industry pushes toward higher levels of autonomy and digital complexity, Michael is diving deep into the frameworks required to ensure these systems are provably safe. What Michael is covering: - Safety Case Assessment: How we partner with commercial customers to build rigorous, evidence-based safety arguments. - LLMs in Safety: Exploring the frontier of Large Language Models and their role and risks in safety-critical applications. - Safety Culture & SPIs: Moving beyond code to measure the human element of safety through actionable Safety Performance Indicators. Drop a comment below or stop by to connect with Michael and the Edge Case team!

More reviews do not solve a visibility problem. They help at a point in time, but they do not keep pace with continuous change. As systems evolve faster, the distance between reviews becomes more significant. Edge Case sees organizations relying on review cycles that were never designed for the systems they are building today.

Safety has traditionally been treated as an engineering and compliance discipline. It is now also a data problem. The challenge is no longer just defining hazards or writing requirements. It is ingesting, organizing, and interpreting massive volumes of evidence in a way that supports real decisions.

There was a time when a team could fully understand a system’s risk profile. That assumption is breaking. With constant software updates, AI behavior shifts, and continuous operational input, the system is always changing. Keeping safety aligned now requires tracking more variables than any team can manually manage.