CalCompute — Ungoverned AI in California's Freight Corridor

“The algorithm made the decision. The community absorbs the cost. No governance layer exists to account for that trade-off.”

A note on direct observation: As Veterans Treatment Court Liaison for the County of Los Angeles Department of Military and Veterans Affairs, the author drives regularly through the Cities of Commerce and Vernon — major freight transportation corridors connecting Long Beach Harbor to inland California. The backed-up trucks, the surface street diversions, the diesel exhaust in working-class neighborhoods: these are not abstractions in a logistics report. They are what the route to Compton Court looks like on a Tuesday morning. This paper is grounded in that observation.

Introduction: The Algorithm Optimizes for the Port

The ports of Los Angeles and Long Beach together handle approximately 40 percent of all U.S. containerized imports. Every container that moves through those ports eventually reaches a truck. Every truck follows a route. And increasingly, those routes are determined by artificial intelligence.

The AI systems optimizing freight movement at the San Pedro Bay complex are real, funded, and expanding. The Port of Los Angeles received an $8 million state grant in 2024 specifically to enhance its Port Optimizer technology — including an AI-powered truck appointment system designed to reduce congestion and improve throughput. This is AI infrastructure operating at the scale of the U.S. economy.

Here is what that AI does not optimize for: the communities the trucks drive through.

When primary freight routes congest — and they do, regularly, on the 710 and the 5 and the surface streets of Commerce and Vernon — routing algorithms divert loads onto alternative corridors. Those corridors run through some of the lowest-income, most pollution-burdened communities in California. No governance layer exists to account for that trade-off.

This is ungoverned AI in transportation. Not a hypothetical future risk. A daily, documented reality on the roads the author drives to reach Compton Court.

1. The Freight Corridor and the Communities It Runs Through

What the Route Actually Looks Like

The freight corridor connecting the Port of Long Beach to inland California passes through a specific geography that matters for this argument. Commerce, Vernon, Wilmington, and Compton are not interchangeable points on a logistics map. They are communities — working-class, predominantly Latino and Black, with documented histories of environmental burden from the port complex and its associated truck traffic.

Los Angeles and Long Beach rank first in the nation for most polluted ozone, fifth for year-round particle pollution, and seventh for short-term particle pollution. Community members along these truck routes have lived with the health consequences of freight movement for decades. Pediatric pulmonologists in the region have documented that diesel particulate matter from port-related trucks lodges in lung tissue, contributing to asthma attacks and chronic respiratory disease in children.

Groups representing approximately 400,000 residents in port-adjacent communities have consistently argued that current clean air plans do not go far enough — and that communities have already waited decades for the health commitments made to them to be fulfilled.

How Ungoverned Routing Compounds the Problem

Port congestion is not a rare event. It is a structural feature of high-volume freight corridors. When terminal queues back up, when highway ramps are at capacity, when the 710 is running slow — AI routing systems do exactly what they are designed to do. They find another path.

That other path runs through Commerce. It runs through Vernon. It diverts onto surface streets in communities that are already absorbing a disproportionate share of regional air pollution. And the algorithm that made that routing decision was optimizing for shipper efficiency, terminal throughput, and delivery time. It was not optimizing for cumulative community health impact, diesel exposure per residential block, or the equity of distributing freight burden across income levels.

It was not designed to. No one asked it to. And no governance layer exists to require it.

2. AI Is Already in the Freight System — Without Governance

The Port Optimizer and What It Does Not Measure

The Port of Los Angeles Port Optimizer is a real system, receiving real public investment. The 2024 state grant specifically funded enhancements including an AI-powered truck appointment system and improved visibility into greenhouse gas emissions. These are meaningful improvements. A truck appointment system reduces terminal queuing, which reduces idling emissions at the port gate.

But the Port Optimizer’s AI does not currently account for:

Cumulative diesel exposure in communities along diverted routes.
Road degradation costs imposed on municipal infrastructure in Commerce and Vernon.
Noise burden per residential block during off-peak diversions.
The equity distribution of freight health burden across income levels.
Whether the efficiency gain at the terminal produces a net community harm downstream.

These are not exotic metrics. They are documented, measurable, and directly relevant to California’s stated commitments on environmental justice and climate equity. They are simply not in the optimization function. That is what ungoverned AI looks like in practice: not malicious, not broken — just incomplete. Optimizing for what it was asked to optimize for, with no constraint layer requiring it to account for what it was not.

The Feedback Loop No One Is Measuring

The ungoverned freight AI feedback loop works like this:

Terminal congestion triggers route diversions by AI scheduling systems.
Diverted trucks move through Commerce, Vernon, and Compton on surface streets not designed for sustained heavy freight.
Diesel emissions accumulate in communities already ranked among California’s most pollution-burdened.
Road damage from heavy truck loads accumulates on municipal streets, with repair costs borne by local governments.
No audit trail connects the routing decision to its community impact.
The same algorithm makes the same trade-off the next day, with no record that it did.

This is not a supply chain problem. It is a governance problem. The algorithm is doing its job. The missing piece is a constraint layer that requires community impact to be part of the job.

3. What Governed AI in Freight Would Actually Do

Governing the Routing Decision

Governed AI in freight logistics does not mean slower delivery or less efficient ports. It means the routing algorithm operates under explicit constraints that include community impact alongside efficiency metrics.

A governed freight routing system would:

Maintain a community impact budget alongside throughput and cost budgets — capping cumulative diesel exposure per residential corridor per day.
Log every routing decision with the variables that drove it, creating an auditable record of when and why freight was diverted through specific communities.
Weight alternative routes that distribute freight burden more equitably, even at a marginal efficiency cost.
Generate a governance report that regulators, community groups, and lawmakers can examine — not as an internal document, but as a public accountability record.

None of this requires a new model. The Port Optimizer already processes real-time traffic data, terminal schedules, and truck appointment windows. Adding community impact as a governed constraint is an architectural decision, not a technical impossibility.

The Efficiency Argument for Governed Freight AI

The case for governed freight AI is not only an equity argument. It is an efficiency argument.

AI-driven route optimization, when properly governed, has documented performance gains: a 50 percent reduction in empty truck miles, 25 percent faster delivery times, and fuel savings in the hundreds of millions of dollars annually at scale. These gains come from better information, not from ignoring community constraints. A governed system that also optimizes for equity is not a less efficient system. It is a more complete one.

California’s freight system handles goods that move through the entire national economy. The AI managing that system should be held to the same standard as every other piece of critical infrastructure in the state: transparent, accountable, and operating within publicly defined constraints.

4. The Community Health Cost of Inaction

The health burden on port-adjacent communities is not speculative. It is documented, measured, and ongoing. The South Coast Air Quality Management District has identified the communities along freight routes in Los Angeles County as among the most disproportionately impacted in the state under California’s environmental justice framework.

What is not yet measured is the specific contribution of AI routing decisions to that burden. When a routing algorithm diverts a column of diesel trucks onto Alameda Street at 2 a.m. because the 710 is backed up, that decision does not appear in any public record. It does not show up in the Port Optimizer’s reported metrics. It does not appear in the Clean Air Action Plan’s progress reports. It happened, it had a measurable health impact, and it was never documented.

Governed AI changes that. Not by preventing all diversions — sometimes there is no alternative. But by requiring that every diversion be logged, attributed to a decision algorithm, and counted against a community impact budget that has public visibility.

What gets measured gets managed. What does not get measured gets repeated.

5. California Already Has the Framework — It Is Missing the Governance Layer

California has passed more environmental justice legislation than any other state. Senate Bill 535 (De León, 2012) requires that 25 percent of California Climate Investment funds benefit disadvantaged communities. Assembly Bill 1550 (Gomez, 2016) strengthened those requirements, mandating that investments be located within disadvantaged communities — not merely directed toward them. Senate Bill 1000 (Leyva, 2016) requires cities and counties with disadvantaged communities to incorporate environmental justice goals into their general plans. Together these laws create a framework that explicitly recognizes that communities like Commerce, Vernon, Wilmington, and Compton bear disproportionate pollution burdens and are entitled to equitable treatment under state law.

The Port of Los Angeles has a Clean Air Action Plan that has reduced port-related emissions significantly since 2005. The state has invested $27 million in port AI technology. The commitment to community health is documented in statute and in policy.

What is missing is a governance layer that connects the AI systems now operating in the freight corridor to the equity commitments already encoded in California law. The Port Optimizer can produce a greenhouse gas emissions estimate. It does not currently produce a community health impact account for routing decisions. That gap is not a technical limitation. It is a governance choice that has not yet been made.

Governed AI in transportation does not require California to rebuild its freight infrastructure. It requires California to hold the AI systems already operating in that infrastructure to the same accountability standards it holds every other actor in the corridor.

A Hypothesis for California Transportation

Governed AI inference, applied to freight routing systems operating in and around the Port of Los Angeles and Port of Long Beach complex, will produce a measurable and documented reduction in the community health burden imposed on lower-income corridor communities — including Commerce, Vernon, Wilmington, and Compton — by requiring routing algorithms to account for cumulative diesel exposure, road degradation, and equity distribution of freight burden alongside throughput and efficiency metrics, compared to ungoverned AI routing currently operating without community impact constraints or public audit accountability.

This hypothesis is testable. The Port Optimizer already collects the data needed to measure it. The South Coast Air Quality Management District already monitors the community health outcomes. California’s environmental justice framework already defines the accountability standard. What is missing is the governance layer that connects them.

CalCompute is positioned to define that standard — not as a restriction on California’s freight economy, but as the infrastructure that makes AI a trustworthy partner in managing the most consequential logistics corridor in the United States.

References

California EPA. (2022). AB 1550: Climate investments for California communities. Assembly Bill 1550 (Gomez, Chapter 369, Statutes of 2016). https://calepa.ca.gov/envjustice/ghginvest

California Office of the Attorney General. (2025). SB 1000: Environmental justice in local land use planning. Senate Bill 1000 (Leyva, Chapter 587, Statutes of 2016). https://oag.ca.gov/environment/sb1000

California Office of Environmental Health Hazard Assessment (OEHHA). (2022). SB 535 disadvantaged communities designation. Senate Bill 535 (De León, Chapter 830, Statutes of 2012). https://oehha.ca.gov/calenviroscreen/sb535

CalMatters. (2025, March 21). ‘Herculean effort’: These port communities have waited decades for clean air. Why a new plan may fall short. https://calmatters.org/environment/2025/03/port-communities-air-pollution-plan-losangeles-long-beach

Cargofive. (2025). The future of AI and machine learning in freight management. https://cargofive.com/future-of-ai-machine-learning-freight-management

Clean Air Action Plan. (2024). San Pedro Bay ports clean air action plan — trucks. https://cleanairactionplan.org/strategies/trucks

DeBacco Nexus LLC. (2026). Empirical research tier catalog: Inference governance module [Internal research documentation]. Patent Pending USPTO 19/571,156. Available upon request.

FreightAmigo. (2025). LA & LB ports congestion: How to navigate. https://www.freightamigo.com/en/blog/logistics/losangeles-and-long-beach-ports-navigating-the-congestion-crisis

Long Beach Business Journal. (2022, March 7). Environmental health hazards impacting the city of Long Beach. https://lbbusinessjournal.com/environmental-health-hazards-impacting-the-city-of-long-beach

Port of Los Angeles. (2024, July 16). Port of Los Angeles awarded $8 million to accelerate its Port Optimizer technology. https://www.portoflosangeles.org/references/2024-news-releases/news_071624_go_biz_grant

Supply Chain Dive. (2024, July 18). Port of Los Angeles to upgrade Port Optimizer after California grant. https://www.supplychaindive.com/news/port-los-angeles-port-optimizer-technology-8-million-grant/721540