Moxion Power: Engineering Battery Systems That Have to Work Every Time

There is a particular kind of engineering pressure that doesn’t come from a standards committee. It comes from a film production shutting down for six hours because the generator failed, or a construction site going dark, or an outdoor event losing power during a performance. No one writes a formal nonconformance report. They just never call you again.

Moxion Power, based in Richmond, California, builds mobile battery-powered energy storage units — essentially, large-format lithium battery systems packaged into ruggedized, towable units that can replace diesel generators on film sets, live events, and construction sites. The company was founded on the premise that the generator replacement market was ready for electrification: that battery energy density, power electronics maturity, and total cost of operation had crossed a threshold where diesel was no longer the default.

That premise is proving out. But the engineering required to make it work at field scale is substantially more demanding than the product category implies.

What Moxion Actually Builds

A Moxion unit is not a large UPS in a trailer. It is a system integrating high-voltage lithium battery packs, a battery management system (BMS), thermal management hardware, power conversion electronics, user-facing control interfaces, telematics, and a mechanical enclosure designed to survive outdoor deployment across climate extremes — then be towed to the next job site and do it again.

The power output specifications needed to serve real production loads — lighting rigs, grip equipment, craft services, HVAC — require high-voltage DC bus architectures with AC conversion stages capable of handling highly variable, non-linear loads. Film sets in particular are notorious for simultaneous high-demand transients: a bank of tungsten fixtures coming on at once, motor-driven equipment cycling, uncoordinated load steps that would trip a lesser system.

Construction sites add a different load character: long duty cycles at moderate-to-high power, dust, vibration from nearby equipment, and operators who are not electricians and have minimal patience for complex interfaces.

Events sit somewhere between the two — shorter deployment windows, often higher peak loads, and a political environment where any visible failure lands on the vendor immediately.

Each of these use cases imposes different requirements on the same physical system. That is a systems engineering problem, not just a product engineering problem.

The Certification Gap — and Why It Matters

Here is where Moxion’s situation becomes genuinely interesting from a systems engineering perspective.

Products in adjacent domains operate under explicit certification regimes. Medical devices have FDA pathways and IEC 60601. Automotive battery systems have ISO 26262 functional safety requirements and UN ECE R100 for EV batteries. Aviation has DO-178C and DO-254. Grid-scale stationary storage has UL 9540 and NFPA 855. Each of these frameworks mandates a structured requirements engineering process as a precondition for deployment.

Moxion’s product category — mobile, temporary, non-utility-connected power equipment — does not have an equivalent mandating framework. There are relevant standards: UL 2580 for EV battery systems, UL 9540 elements that apply if you want fire safety listings, NFPA 70 (NEC) for electrical connections, OSHA regulations for construction site electrical. But none of these together constitute a comprehensive functional safety mandate that forces a full hazard analysis and requirements traceability chain.

This creates an underappreciated engineering governance challenge. When a standard mandates that you perform a Failure Modes and Effects Analysis (FMEA) and trace every safety-relevant requirement to a verification record, the standard does the organizational forcing function for you. When there is no such mandate, an engineering team has to impose that discipline on itself — and sustain it under schedule pressure, resource constraints, and the perpetual temptation to ship.

Self-imposed rigor is harder than mandated rigor. Anyone who has worked in a startup that later pursued aerospace or medical certification knows this firsthand: the retroactive cost of building a safety case for a system that was engineered without one is enormous.

Three Domains, One Box

Moxion’s core systems engineering challenge is that their product must simultaneously satisfy constraints from three engineering domains that each carry their own failure modes, their own design vocabulary, and their own set of things that kill you if you get them wrong.

High-voltage electrical systems carry obvious hazards: arc flash, electric shock, thermal runaway propagation through bus faults, ground faults in wet outdoor environments. The BMS must monitor cell state at sufficient resolution and speed to catch developing faults before they become field events. Protection coordination — the hierarchy of fuses, contactors, and software limits — has to be analyzed for every fault path, not just the obvious ones. In a mobile system deployed by non-specialist operators, the assumption that someone will intervene intelligently in a fault condition is dangerous.

Thermal management is where lithium battery system design most commonly breaks down at scale. Cell performance, cycle life, and — critically — the boundary conditions for thermal runaway are all temperature-dependent. A system that operates fine in a 65°F Richmond morning can behave very differently on a Las Vegas film location in August, or during a winter outdoor event in the Mountain West. Active thermal management — liquid cooling or forced air with intelligent control — adds complexity, cost, and its own failure modes. Passive thermal management is simpler but limits the operating envelope. The design choice here cascades into system-level requirements for ambient temperature range, derating behavior, and what the system does autonomously when it’s being pushed beyond its thermal comfort zone.

Field durability is the third domain, and it is often the one that surprises engineering teams. Vibration profiles from road transport are harsh and broad-spectrum. Connectors that work perfectly in a lab fail after six months of daily mating cycles by operators wearing gloves. Enclosure sealing that passes ingress protection testing degrades with UV exposure and thermal cycling. Telematics that work on a San Francisco production lot may have intermittent connectivity on a remote construction site, which means the system has to fail gracefully when cloud connectivity is unavailable.

The failure mode that ends a company in this market is not a dramatic event — it is accumulated nuisance failures that erode customer trust until the rental house goes back to diesel.

Requirements Without a Framework Forcing Function

How does a company like Moxion define requirements rigorously when no external authority is requiring them to?

The honest answer is: with difficulty, and with varying degrees of success across the industry. Some companies in this space treat their product as consumer electronics with higher stakes. Others approach it closer to industrial equipment, with structured hazard analysis and explicit safety requirements that get tracked to verification evidence.

The engineering demands of the product itself push toward rigor. When you are designing a system that stores megajoules of electrochemical energy, sits outdoors in variable weather, and gets connected and disconnected by operators of unknown electrical sophistication, the consequences of underspecified requirements show up quickly and visibly.

The thermal runaway question alone — what is the required behavior of the system when a cell enters thermal runaway? What does the BMS do? What does the enclosure need to contain? What is the expected outcome for the operator and bystanders? — generates a chain of functional requirements, design constraints, and verification tests that can fill a requirements management system if you’re working the problem correctly.

The field reliability question generates another chain. Mean time between failures for field-deployed units translates into maintenance interval requirements, which translate into component selection criteria and acceptance test requirements for incoming parts. If you’re not tracking these as explicit requirements with verification records, you’re hoping rather than engineering.

There’s no ISO 26262 to mandate an Automotive Safety Integrity Level analysis here. But the physics of the product creates its own forcing function — eventually. The question is whether the engineering team gets ahead of it or waits for field events to drive the learning.

The Market Pressure Dynamic

One factor that complicates the engineering discipline picture is market pressure from the customer side. Film productions operate on aggressive schedules with brutal cost sensitivity. Rental houses — the intermediary channel through which most production companies access power equipment — make purchasing decisions based on day rate economics and reliability reputation, not on the depth of a vendor’s hazard analysis.

This creates a dynamic where engineering rigor has to be justified internally on the basis of long-term reliability and safety outcomes, not on immediate commercial requirements from buyers. The buyer doesn’t ask for the FMEA. The buyer asks whether the unit will show up charged and work all day.

That buyer expectation is, in a roundabout way, a very stringent reliability requirement. It just isn’t written down anywhere that the vendor is legally required to address it.

The companies that will win this market long-term are the ones that internalize the discipline anyway — that build structured requirements processes not because an auditor is checking, but because the alternative is field failures that end customer relationships and, in a worst-case scenario, create safety events that are difficult to survive reputationally and legally.

Honest Assessment

Moxion is operating in a genuinely difficult engineering space and doing so at a time when the tools available for managing systems complexity — AI-assisted requirements analysis, graph-based traceability, connected digital threads — are better than they have ever been, even if adoption in this segment of the industry lags behind aerospace and automotive.

The generator replacement market is real. The environmental and operational case for mobile battery storage over diesel is strong and getting stronger as battery costs continue their long-term decline. The regulatory environment, while not yet mandating comprehensive safety cases for this product category, is moving in a direction that will eventually close that gap — particularly if there are significant field safety events from any competitor in the space.

The engineering challenge Moxion faces is not exotic. It is the well-understood problem of integrating multiple high-consequence subsystems into a reliable field product, under cost and schedule pressure, without the external forcing function that regulated industries use to maintain discipline. The companies that treat that challenge with appropriate seriousness — that build the requirements infrastructure even when no one is requiring them to — will be well-positioned when the market matures and the bar rises.

The companies that treat it as a software startup with a hardware attachment will find the learning curve expensive.