Safer, Faster Public Transportation: AC Transit’s AI-Powered Upgrade with Hayden AI

NVIDIA AI Podcast – Episode 290
Date: February 18, 2026
Host: Noah Kravitz
Guests: Hasan Beg (Chief Technology Officer, AC Transit) & Marty Beard (CEO, Hayden AI)

Overview

In this episode, NVIDIA AI Podcast host Noah Kravitz explores how AC Transit and Hayden AI are collaborating to transform public transportation in California’s East Bay through advanced AI and computer vision technology. The conversation centers on deploying AI-powered automated bus lane and stop enforcement to improve transit speed, safety, and accessibility, all while addressing public concerns about privacy and transparency.

Key Discussion Points & Insights

1. Introduction to AC Transit and Hayden AI

• AC Transit: Third largest bus operator in California, covering Alameda and Contra Costa counties. Handles 55–57 million annual riders pre-COVID, with a core mission focused on safety, security, reliability, and sustainability. Notably, it has an elected Board, which is rare among transit agencies.

  • (Hasan Beg, 01:27):

    "Our mission at AC Transit is safe, secure, reliable in sustainable public transit... Board members, elected board members are... passionate about public transit and... services."

• Hayden AI: Founded in 2019, a San Francisco-based company specializing in AI-powered transit solutions. They equip over 2,100 vehicles in 10+ US cities (and expanding internationally), focusing on improving transit speed and safety.

  • (Marty Beard, 04:09):

    "At this point, we're on over 2,100 vehicles nationally and working in 10 major cities across the country."

2. Genesis of the Collaboration

• The Challenge: AC Transit struggled with illegally parked cars in dedicated bus lanes, leading to delays and reduced efficiency. Their manual, legacy enforcement (requiring drivers to capture images and submit evidence) had a less than 5% citation success rate and placed extra burden on bus operators.

• Solution Sought: An automated system that could streamline and improve enforcement accuracy and operator experience.

  • (Hasan Beg, 06:06):

    "...our success rate was less than 5%... So that was a major business problem... We worked with our partners in crafting the new legislation, which is AB917, that authorizes us to use... automated lane enforcement technology."

• Pilot Program: Started with five buses, Hayden AI was identified as the best partner. The program expanded after success in the pilot phase.

  • (Hasan Beg, 07:07):

    "We started the whole journey starting with five buses and a pilot. ...Now we have been working for almost more than two years now."

3. The Technology and Process

• System Architecture:

  • Hardware: Cameras installed inside the bus, facing outward to monitor bus lanes and stops.

  • Edge Computing: Images processed in real time on on-board control boxes using NVIDIA’s edge AI products.

  • Computer Vision: Detects vehicles illegally parked in bus zones or stops, integrating object recognition and location data (“sensor fusion”).

    • (Marty Beard, 09:50):

      "You've got this control, that's where the AI is running and that's looking for the violation. ...That's optimized for that."

  • Privacy by Design: No images or personal info about passengers are collected or stored; only license plates of vehicles in violation are captured.

  • Enforcement Workflow: The AI system submits probable violations to authorities for human review before tickets are issued.

    • (Marty Beard, 11:46):

      "We package what we believe is a violation... But that then does get sent for... ultimate review by somebody..."

4. Impact – Operator & Public Feedback, Metrics of Success

• For Bus Operators:
  • Automation removes the cognitive and physical demands of identifying and capturing violations while driving.
    • (Hasan Beg, 12:28):

      "From the operator's perspective... it's a big blessing that they don't keep pressing the button."

• System Results:
  • On-time performance and accessibility improved.
  • 70% reduction in first-time offenses (illegally parked vehicles).
  • Enhanced ability for buses to pull in and for accessibility boarding.
  • Higher accuracy in violation data, making enforcement fairer and more reliable.
    • (Hasan Beg, 13:32):

      "We are seeing reduction 70%. ...We are seeing improvements... on time performance... accessibility..."

5. Legislation, Privacy, & Public Education

• Stakeholder Communication:
  • Transparent engagement with the public, legislative bodies, and operators about objectives and privacy.
  • Compliance with strict privacy policies and state legislation (AB917), which allows and regulates automated enforcement.
  • Legislation is subject to review and renewal; ongoing data collection and reporting is required to justify continued use.

    • (Marty Beard, 15:01):

      "When you talk about AI and cameras... it's kind of like, okay, yeah, just step back for a second. This is not looking at people..."

    • (Hasan Beg, 16:39):

      "...we educated them, we presented to them... specific to like a privacy and we wanted to make sure that we are in compliance..."

6. Broader Digital Transformation & Future Directions

• For AC Transit:

  • Emphasis on innovation, pilot-testing, and using leading-edge technology to directly address core business problems.
  • Support from an elected Board and executive team aids in transformative projects.
    • (Hasan Beg, 21:15):

      "I don't believe in deploying any technology... if it is not solving my business problem."

• For Hayden AI:

  • Expansion into related use cases: bike lane enforcement, parking management, roadworks (construction zone) identification.
  • Vision of “practical AI” as a tool to help city managers improve transit, infrastructure, and overall urban mobility.
    • (Marty Beard, 22:07):

      "We talk a lot here about practical AI. ...We can help manage assets. ...buses, bikes, parking vehicles..."

7. Advice to Other Agencies

• For Public Sector CTOs:

  • Start by focusing on clear business problems.
  • Partner closely with internal and external stakeholders.
  • Pilot and iterate—be willing to “fail fast” and learn quickly.
  • Share best practices openly with the public sector community.
    • (Hasan Beg, 25:06):

      "If I have success, if I have good methodology, good finding, and the way to make it happen, you know, we all share. ...Just be bold, just to try out, you know, and really shoulder to shoulder with the business."

• From the Tech Vendor Perspective:

  • Drive adoption by demonstrating practical results, not just theory.
  • Use data to illustrate real-world benefits and build confidence among agencies.
    • (Marty Beard, 26:37):

      "What is the business problem you're trying to solve? ...So I think it's fun to go out and quote, sell the vision because, you know, it can work."

Memorable Quotes & Timestamps

• On the Challenge Facing AC Transit:

  • Hasan Beg, [06:06]:

    "Our success rate was less than 5%. So that was a major business problem."

• On Privacy Concerns:

  • Marty Beard, [15:01]:

    "When you talk about AI and cameras... it's kind of like, okay, yeah, just step back for a second. This is not looking at people. There are no people identified. Right. This is only looking at vehicles..."

• On Organizational Culture for Innovation:

  • Hasan Beg, [21:15]:

    "I don't believe in deploying any technology or a technical solution if it is not solving my business problem."

• On “Practical AI” and Future Applications:

  • Marty Beard, [22:07]:

    "We talk a lot here about practical AI... It's a fantastic place where new tech like AI is kind of meeting reality."

Important Segment Timestamps

• [01:27] — Hasan Beg introduces AC Transit and its unique structure and mission.
• [04:09] — Marty Beard on Hayden AI's reach and expertise.
• [06:06] — Description of legacy enforcement challenges at AC Transit.
• [09:23] — Marty Beard outlines the technical architecture of the enforcement system.
• [13:13] — Early impact: operator relief and metrics from system deployment.
• [15:01] — Addressing privacy and public perception around AI and cameras.
• [21:15] — The importance of connecting technology to business outcomes.
• [22:07] — Broader role of AI in public transit and city management.
• [25:06] — Advice for public-sector leaders: share, experiment, focus on business need.

Resources

• AC Transit: actransit.org
• Hayden AI: hayden.ai (Also active on LinkedIn)

Tone & Style

The episode was practical and optimistic, balancing real-world operational realities with an inspiring look at how technology—when thoughtfully designed and implemented—can change public systems for the better. The speakers emphasized transparency, public benefit, collaboration, and continual improvement, using straightforward, accessible language with a clear focus on solving real community problems.