Winning the Room: Voxel to Lead Critical AI Adoption Panel at ASSP Safety 2026

The push for artificial intelligence in the industrial sector often feels like a "forced march" for the frontline workers actually wearing the high-vis vests. At the upcoming ASSP Safety 2026 conference in Anaheim, computer vision pioneer Voxel is looking to flip that script. Brittany DeRafelo, Voxel’s Director of Customer Solutions, is set to moderate a high-stakes panel titled "Working Smarter, Not Harder: Embracing AI Without Forcing Adoption." It's a move that signals a shift away from "tech for tech's sake" and toward a more human-centric strategy for site safety.

The session, scheduled for the American Society of Safety Professionals (ASSP) annual gathering from June 15-17, isn't just a corporate showcase. Instead, it features a heavy-hitting lineup of risk and safety veterans who’ve been in the trenches. The panel includes Lisa Battino from United States Cold Storage, Macy’s Terri Rybacki, and Shane Quinn of MS International. These leaders aren't there to pitch software; they’re there to discuss the messy, real-world friction of introducing AI-powered site visibility platforms into environments where "trust" is a harder metric to move than incident rates.

Voxel’s core technology is built on a clever premise: stop asking companies to buy new hardware and start making their existing cameras smarter. By leveraging computer vision to identify ergonomic risks or vehicle near-misses in real-time, the platform provides what Business Wire describes as actionable insights that shift safety from reactive to proactive. As DeRafelo pointed out in the lead-up to the event, the biggest barrier to AI isn’t the code—it’s the culture. If teams feel the technology is being "done to them" rather than "for them," the most advanced algorithm in the world won’t save a single soul on the warehouse floor.

Breaking the Barrier of "Tech Fatigue"

The conversation in Anaheim is expected to dive deep into the psychology of buy-in. Safety professionals frequently deal with "tech fatigue," where new systems are viewed as just another layer of surveillance or administrative burden. The panel aims to dismantle this perception by showing how AI can act as an assistive tool that validates human expertise rather than replacing it. By focusing on privacy-first designs and clear, non-punitive applications, these industry leaders hope to provide a roadmap for digital transformation that employees actually want to follow.

Why Trust Trumps Tools

Success in this space is increasingly measured by how well a tool integrates into the daily flow of a site manager's life. Voxel has seen significant momentum recently, following a $44 million Series B round, by leaning into this philosophy. The panel at Safety 2026 will likely serve as a litmus test for the broader industry: can we move past the hype and prove that AI is a partner in the safety journey? For the thousands of safety professionals descending on the Anaheim Convention Center, the answer to that question will define the next decade of workplace protection.

The Hidden Friction of Industrial Automation

Beyond the Spec Sheet: The real battle for AI in the warehouse isn’t fought in the C-suite; it’s won or lost on the loading dock during a graveyard shift. While tech companies often tout "seamless integration," seasoned safety managers know that any new camera-based system is immediately viewed through the lens of "Big Brother." Voxel’s decision to headline a panel specifically on not forcing adoption acknowledges a hard truth that many Silicon Valley startups ignore: industrial workers are historically cynical about tools that feel like automated supervisors. By bringing names like Macy’s and MS International to the table, the discussion shifts from theoretical efficiency to the gritty reality of earning a forklift operator’s trust.

Historical context tells us that safety technology has often been a double-edged sword. For decades, "safety" meant more paperwork and more disciplinary triggers, which created a culture of concealment where near-misses were swept under the rug to avoid penalties. Voxel is attempting to break this cycle by repositioning computer vision as a passive guardian rather than an active snitch. The nuance here lies in the data loops; if the AI identifies a spill or a blocked fire exit, the value isn't in punishing the person who left it there, but in identifying the systemic bottleneck that led to the hazard. This distinction between "surveillance" and "visibility" is the cornerstone of the panel's argument for organic adoption.

Stakeholder perspectives within these organizations are often at odds, which is why the inclusion of United States Cold Storage is particularly telling. In temperature-controlled environments, worker fatigue and physical strain are amplified, making ergonomic AI an easier "sell" to the workforce if it’s framed as a health benefit. When Brittany DeRafelo leads this conversation, she is essentially acting as a bridge between the data-hungry insurers—who want lower premiums through risk mitigation—and the site leads who just want their teams to get home in one piece. Balancing these conflicting interests requires more than just a robust API; it requires a transparent communication strategy that explains exactly where the data goes and, more importantly, where it doesn't.

The technical shift toward using existing infrastructure is also a clever psychological play. There is a specific kind of anxiety that comes with seeing new, high-tech hardware bolted to the ceiling. By turning the "dumb" CCTV cameras that have been there for years into "smart" sensors, the physical environment remains familiar to the employees. This reduces the "newness" friction and allows the conversation to focus on the outcomes—like fewer collisions and reduced muscle strain—rather than the alien nature of the equipment itself. It’s a strategy of stealthy improvement that prioritizes the user experience over the novelty of the gadgetry.

Ultimately, the ASSP Safety 2026 panel serves as a microcosm for the broader "Industry 4.0" movement. The era of mandates is ending because, quite simply, mandates don't work in high-turnover environments where morale is a fragile resource. The industry leaders joining Voxel are betting that the future of safety isn't found in a thicker rulebook, but in a smarter pair of eyes that helps workers navigate their day with less risk. Success will be defined by whether a worker sees a Voxel-enabled camera and thinks of it as a safety net rather than a stopwatch, marking a permanent change in how technology is introduced to the American workforce.

The Paradox of Passive Oversight

Reading Between the Lines: There is a persistent irony in the marketing of "invisible" AI safety tools: the more seamless the technology becomes, the more it risks becoming an unaccountable black box. While Voxel champions the idea of "embracing without forcing," there is an inherent contradiction in using existing security cameras to monitor workforce behavior. Even if the intent is purely for ergonomic safety or spill detection, the infrastructure remains the same as that used for loss prevention and disciplinary oversight. The challenge for the ASSP panel will be convincing a skeptical labor force that the "smart" layer added to these cameras doesn't just provide a more sophisticated way to micromanage the seconds of a worker's day under the guise of their own well-being.

Measured skepticism suggests that "organic adoption" is often a euphemism for a more palatable form of inevitability. When major players like Macy’s and United States Cold Storage implement these systems, the choice for the individual worker often disappears, regardless of how much the technology is "sold" rather than "mandated." If the AI begins to dictate the pace of work by flagging every deviation from a "perfect" ergonomic posture, it may inadvertently increase cognitive load and stress. This creates a feedback loop where the drive for physical safety could potentially degrade psychological safety—a metric that is far harder for computer vision to track than a forklift’s speed or a worker’s lumbar curve.

The projection for 2026 and beyond indicates that the "human-centric" approach will live or die by the transparency of its data usage. The industry currently lacks a universal standard for AI data governance in the workplace, leaving a vacuum that corporate policies fill on a case-by-case basis. If Voxel and its peers cannot guarantee that "safety data" remains strictly siloed from "performance data," the trust they are working so hard to build will evaporate at the first sign of a productivity-based firing triggered by an algorithm. The implication is clear: the tech is the easy part, but building a legal and ethical framework that protects the worker as much as the warehouse is the real hurdle.

Furthermore, we must consider the risk of "algorithmic complacency," where site managers begin to over-rely on the software to catch hazards, potentially dulling their own situational awareness. If the dashboard doesn't flash red, is the floor truly safe? The danger of AI in safety is that it might move the responsibility of vigilance from the human eye to a server rack, creating a false sense of security in the gaps where the camera cannot see. For the safety industry to truly evolve, these panels must address not just the success stories, but the potential for these tools to inadvertently create new, unforeseen risks in the workflow.

In the end, the greatest triumph of industrial AI might just be convincing a forklift driver that a camera is his new best friend, though he'll likely still suspect it’s the only friend he has that’s secretly reporting his posture to an insurance adjuster in New Jersey.

Artūras Malašauskas is an AI Systems Integrator with 20+ years of production-grade web engineering experience. He has designed, shipped, and scaled enterprise Python/PHP systems for logistics, SaaS, and public-sector clients. For the past year, he has focused exclusively on AI integrations: deploying open-source LLMs, building generative media pipelines (image, audio, video), and engineering multi-agent workflows for real production environments. His standard: reproducibility, security, cost-efficient inference—no vaporware. He documents and evaluates emerging AI tooling, separating verified capabilities from marketing noise. Technical editor at: muza-ai.eu, ai-verslas.lt, ai-naujinos.lt Connect on LinkedIn