AVO Zone Challenge FAQs
The following questions were submitted during the AVO Zone Challenge informational webinar and through email. Responses have been edited for clarity and brevity. In the event of any discrepancy, the official Participant Package and Challenge documentation governs. Proposals are due on July 16, 2026.
Jump to the question category by clicking below:
1. Autonomous Vehicle Detection & Scope
Q1.1: For the autonomous vehicles you are seeking to monitor, will they be clearly identifiable in the field? For example, will they be branded vehicles such as Waymos, or vehicles equipped with visible sensor suites and AV hardware? Or should participants assume that some target vehicles may appear similar to conventional vehicles?
Will any reference materials be available to support identification or validation of known AV test-fleet vehicles, such as sample images, vehicle types, fleet descriptions, videos, 3D scan data, or other approved non-sensitive materials?
Answer:
At this time, autonomous vehicles (AVs) operating in the District are generally identifiable through visible sensor suites, external branding, or both. Current operators, including Waymo and Zoox, utilize vehicles equipped with recognizable sensors and markings.
Participants may assume that identification will primarily rely on visible characteristics. DDOT has a collaborative relationship with AV operators and may request additional identification materials in support of the AVO Challenge project. At this time, no dedicated reference materials are available.
DDOT is also aware of licensed test vehicles operating in the District and may be able to assist with validation. While license plate recognition may be used as a secondary verification method, DDOT does not intend for license plate readers to serve as the primary means of AV identification.
Q1.2 For purposes of the pilot, will DDOT or the project team identify the AV fleet operator(s) or vehicle types expected to operate in the observation zones, or should applicants propose a validation process that can be adapted after award?
DDOT will be able to provide a list of the AV fleet operators currently conducting testing in the District. Active testing entities are also listed on our DDOT AV webpage.
Q1.3: This is focused on AV identification and behavior detection; is this because there is already a solution/data being collected to track the behavior of regular car/bike/pedestrian traffic?
Answer:
No. DDOT does not currently have a comprehensive system to collect and analyze trajectory, near-miss, and behavioral data for all roadway users.
The vision for this challenge is to capture both AV behavior and the broader transportation context, including interactions involving pedestrians, cyclists, transit vehicles, and conventional vehicles. Understanding how AV behavior compares to human-driven vehicle behavior is an important component of the research effort.
Q1.4: Any possibility of mandating the operating AV vehicles to share their sensor fusion dataset? Including object trajectories.
Answer:
No. DDOT cannot guarantee access to proprietary AV sensor fusion data or trajectory datasets.
DDOT is actively engaging with AV operators and may explore opportunities for collaboration; however, participants should not assume that vehicle operators will provide access to proprietary data as part of this challenge.
Q1.5: When you refer to autonomous vehicles, are you focused solely on road-going passenger vehicles, or should we also consider other autonomous systems operating in the public realm, such as sidewalk delivery robots, autonomous shuttles, or other robotic mobility devices?
Answer:
The primary focus of the challenge is on-road autonomous vehicles.
This includes passenger vehicles and could potentially include autonomous freight or transit vehicles operating on public roadways. Sidewalk delivery robots and similar devices are not the focus of this effort, although they may be present within the broader transportation environment.
Q1.6 Will DDOT, GW, UW, or US Ignite participate in a sample review process for candidate AV detections, multimodal behavior detections, or safety-relevant events? If yes, should applicants propose a validation sample size and review cadence, or will those be established after award?
Answer:
DDOT, GW, and UW will provide feedback to the Challenge Winner on the quality of the solution results. Applicants are encouraged to propose a validation method and should expect feedback on the approach prior to the start of the pilot.
2. Technology, Infrastructure & Deployment
Q2.1 Will this be observed from already established traffic cameras, or will there need to be hardware installed in specific areas?
Answer:
Participants should assume that additional hardware will likely need to be installed as part of their proposed solution.
While DDOT maintains CCTV cameras throughout the transportation network, these cameras are not recorded and are primarily pan-tilt-zoom devices rather than fixed-position monitoring systems. Due to cybersecurity and infrastructure access considerations, DDOT generally prefers solutions that deploy independent hardware rather than integrating directly with existing traffic camera infrastructure.
However, proposals that leverage existing CCTV assets may be considered and discussed further during implementation planning.
Q2.2 The Participant Package notes that two Ouster LiDAR sensors are installed at New Jersey Avenue SE and M Street SE and may be available to the Challenge Winner. What access methods, data formats, documentation, and technical interfaces should applicants assume for those sensors?
The processed data is currently only available through the BlueCity portal. There are plans to make this available on the edge device so it can be accessed directly there in early 2027, which may make it available in time for this project.
DDOT can download the data from the portal in Excel format. Trajectory data files for out-of-crosswalk events (OCW) can be downloaded as a text file, but everything else will be in Excel format (.xlsx).
Q2.3 Because the package identifies existing Ouster LiDAR at New Jersey Avenue SE and M Street SE, should applicants assume they must provide their own sensing approach for 4th Street SW and M Street SW, subject to DDOT approval and installation requirements?
Answer:
Yes, applicants must provide a sensing solution in the absence of existing DDOT infrastructure.
Q2.4 Can applicant-provided temporary edge devices, cameras, LiDAR units, or other sensors be mounted on DDOT-approved vertical signal poles or combination signal/streetlight poles if installed by DDOT staff or designated contractors? Are there size, weight, power, clearance, or enclosure constraints applicants should assume?
Answer:
Yes, Challenge Applicants can propose vendor-provided equipment as part of their solution. The size, weight, power, clearance, or enclosure requirements of the proposed equipment should be provided in the proposal. DDOT will work with the Challenge Winner to resolve any compatibility issues.
Q2.5 Can you provide more information regarding the signal upgrades, including controller upgrades and other technology being deployed?
During the Challenge Information Session, it was mentioned that the traffic signals at the two intersections within the AVO Zone limits will be upgraded to Advanced Traffic Controllers (ATCs). Could you provide additional information on these planned upgrades, including the anticipated timeline and any other technology being deployed at these intersections as part of this effort?
Answer:
DDOT is currently upgrading signal infrastructure at the project intersections. Existing systems are being replaced with Advanced Traffic Controllers (ATCs), enabling improved traffic management capabilities and access to signal phase and timing information.
The expectation is that participants will be able to receive one-way access to signal-phase and timing data. Solutions should not assume the ability to send commands or information back to DDOT signal controllers.
The signal at M St and New Jersey Ave SE is being upgraded this summer to support the M St Reverse Pitch project. The signal at M St and 4th St SW will be upgraded this fall as part of the first phase of the citywide upgrade. No other technology is being deployed alongside this upgrade. The M Street Reverse Pitch is deploying several technologies along the corridor, but proposers should not assume that any of that technology will still be in place by the time this project commences.
Q2.6 Will signal phase and timing data, controller event logs, or other signal-related data be available to the Challenge Winner during the pilot? If yes, what format, cadence, and access method should applicants assume?
Answer:
DDOT can provide documentation for traffic signal timing and phasing. Other data may be available, but DDOT will need to understand the vendor’s need for it (if the data exists). Please note that these signals currently operate on fixed time (set cycles) with no detection, and this is not expected to change at this time, even with the signal upgrades.
Q2.7 What are the inputs in visual understanding here exactly? Are we scoping only traffic cameras, or will other inputs, like dash cameras from other vehicles, be included as well?
Answer:
The primary focus is on fixed infrastructure-based observation systems capable of monitoring traffic activity from a consistent vantage point.
While participants may propose alternative approaches, DDOT’s preference is for infrastructure-mounted solutions that provide systematic, repeatable observations over time.
Q2.8 In addition, you mentioned multimodal understanding. Are we talking about traffic logs like sensors that also output text/audio?
Answer:
In this context, “multimodal” refers to multiple modes of transportation, including pedestrians, bicycles, transit vehicles, passenger vehicles, and freight vehicles.
The term does not refer to audio, text, or language-based inputs.
Q2.9 Thanks for answering my earlier questions. Given that we are not installing the hardware ourselves on this project, what skills or expertise are you looking for in this challenge?
Answer:
DDOT is seeking teams with expertise in sensing technologies, computer vision, AV detection, traffic analytics, data processing, and system operations.
While DDOT will assist with installation on its infrastructure—including access to poles, traffic cabinets, and power connections—the selected team will be responsible for providing and managing the equipment, software, data collection, processing capabilities, and overall operation of the solution.
Q2.10 Any restrictions on sensor/computing vendors, like Chinese hardware normally prohibited under FAR section 889 (telecom hardware)?
Answer:
While this pilot is not a direct procurement of permanent infrastructure, DDOT is interested in solutions that could potentially be deployed more broadly in the future.
Applicants should consider federal compliance requirements and, where applicable, provide a roadmap toward compliance with relevant procurement and technology standards.
3. Installation & Operations
Q3.1 Who will we be coordinating with to install cameras and potentially pull permits if needed?
Answer:
DDOT will coordinate installation activities with the selected team.
Participants should not obtain permits or install equipment on DDOT infrastructure independently. DDOT staff and approved contractors will work with the selected vendor to facilitate installation on traffic poles, cabinets, and other transportation infrastructure as necessary.
Q3.2 This work will require servers, CCTV, and monitoring for how many hours per day?
Answer:
DDOT has not established a specific hourly monitoring requirement.
The amount of data collection required will depend on the frequency of AV activity within the observation zones. Applicants should propose an approach that can effectively capture relevant AV observations while supporting the challenge objectives.
Q3.3 Will proposing the optional straight-road segment along M Street be scored as added value, or is it considered supplemental and non-essential relative to the two required intersections? If the $50,000 award is insufficient to cover the optional segment, should applicants describe it as a future option or a cost-share option?
Answer:
Proposing a solution for the optional straight-road segment will not directly contribute to the evaluation criteria. Proposers are encouraged to describe their capabilities along the corridor to inform DDOT of those capabilities for use in the Challenge Pilot or future planning.
4. Data Collection, Privacy & Research Objectives
Q4.1 What sort of incidents or behaviors are we looking for? Are there evaluation criteria we should consider from the Autonomous Vehicles?
Answer:
The challenge is not focused on enforcement. Rather, DDOT seeks to understand how autonomous vehicles operate within real-world urban environments and how they interact with other roadway users.
The goal is to observe operational patterns, behaviors, and interactions, and to assess potential impacts on the transportation system. Participants should refer to the Technical Requirements section of the Participant Package for examples of behaviors and performance measures of interest.
Q4.2 Are there any goals around privacy, or is the goal to have a live video feed similar to CCTV?
Are there specific privacy, retention, access, or deletion limits applicants should assume for short validation clips, still images, or license plate information used only for supporting validation?
Answer:
Privacy is an important consideration for this project.
DDOT is not seeking a continuously accessible live video feed. Instead, the preference is for solutions that minimize unnecessary data collection, leverage edge processing where possible, and capture only the information necessary to support the research objectives.
Applicants are encouraged to describe privacy protections, data-handling procedures, and any anonymization or filtering approaches in their proposed solution.
Q4.3 If you were to summarize the total output/data from this challenge, what insight are you hoping to capture from it?
Answer:
DDOT seeks to better understand how autonomous vehicles operate in urban environments, how their behavior compares with human-driven vehicles, and how those behaviors affect transportation system performance.
The challenge is intended to provide data-driven insights that support future policy development, operational decision-making, public transparency, and AV oversight activities.
Q4.4 What delivery cadence does GW prefer for raw, near-raw, processed, trajectory-level, object-level, and event-level datasets: daily, weekly, monthly, on-demand, or another schedule?
GW prefers delivery schedules that align with both data fidelity and research needs.
- Raw data delivery is preferred and may be considered essential (at least in the monitoring location of interest and for a given duration exceeding 10 seconds for validation purposes). This includes unprocessed data such as video feeds, sensor outputs, and other source data collected directly from the field.
- High-resolution trajectory-level data is also preferred (depending on the application tested); for example, for some safety applications, a preferred sampling rate would be 10 Hz or higher and decimeter-level spatial resolution or better to ensure sufficient temporal and spatial fidelity for detailed behavioral analysis.
- Processed and object-level datasets are preferred when available, as they improve usability and reduce preprocessing effort for downstream analysis.
- Event-level data are preferred when available, especially for applications involving traffic conflicts, incidents, safety analysis, or behavior-based studies. This can be delivered periodically or bundled with processed datasets.
In addition to scheduled data delivery, GW may request further data transfers, additional analyses, clarifications, or supporting documentation as needed throughout the project lifecycle.
Q4.5 Are there preferred schemas, metadata templates, coordinate systems, file naming conventions, or data dictionaries that applicants should use to support the transfer of data to the GW data hub?
GW prefers datasets to follow standardized schemas and metadata structures to ensure seamless integration into the GW data hub.
Preferred trajectory data schema: Trajectory datasets should include, at a minimum, the following fields:
- ID – Unique identifier for each tracked object
- Time – Timestamp for each observation
- X, Y – Spatial coordinates
- Lane (if applicable) – Lane identifier
- Road (if applicable) – Road segment identifier
- Speed – Instantaneous speed
- Acceleration – Instantaneous acceleration
- Length – Object length
- Width – Object width
- Type – Agent classification
Preferred coordinate systems: Spatial data should use one of the following coordinate reference systems:
- World Geodetic System (WGS 84)
- Universal Transverse Mercator coordinate system (UTM)
These coordinate systems ensure compatibility with mapping and GIS workflows.
Preferred agent classes: Datasets should support the classification of the following agent types:
- Person / Pedestrian
- Bicycle
- Scooter
- Passenger Car
- Autonomous Vehicle
- Motorcycle
- Bus
- Truck (Heavy Vehicle)
Supporting map data: All submitted datasets should be supplemented by map data describing the study area in sufficient detail. This can be done by GW, but in coordination with the participants; the map data should include:
- Road network geometry
- Lane boundaries and lane IDs
- Intersections and traffic control elements
- Relevant study area boundaries or zones
This map layer is important for providing context for trajectory and event-level analysis and for supporting accurate interpretation of vehicle and pedestrian movement patterns.
Q4.6 Should applicants assume responsibility for collecting or integrating weather, lighting, roadway characteristics, and other contextual data, or will DDOT, GW, or UW provide some of these datasets?
Answer:
The Challenge Applicant is primarily responsible for these datasets.
DDOT maintains a robust set of open data (see: opendata.dc.gov), including information on our roadway features:https://opendata.dc.gov/pages/roadway-centerlines. If there are specific agency datasets the vendor would want to access that are not on Open Data DC, DDOT is open to discussing which datasets those are and what access would be needed.
Q4.7 Will the Challenge Winner be permitted to provide public-facing materials such as summary dashboards, infographics, public-safe visualizations, or plain-language findings briefs after DDOT review and approval? If yes, are there preferred formats or review procedures?
Answer:
Community engagement and feedback are planned during the observation pilot period, and the Challenge Winner will be asked to support those engagements. The extent of that support will be developed in the pilot planning phase. DDOT will review any public-facing materials. The format and review procedures will be developed with the Challenge Winner.
Q4.8 Does DDOT have a preferred acceptance-testing process for commissioning, calibration, data quality review, uptime verification, and validation of required outputs, or should applicants propose their own acceptance-testing approach?
Answer:
DDOT, GW, and UW will provide feedback to the Challenge Winner on the quality of the solution results and acceptance testing. Applicants are encouraged to propose a validation method and should expect feedback on the approach prior to the start of the pilot.
Q4.9 Are there any restrictions or requirements on the AI/ML model providers or licensing models (open-source vs. proprietary/closed-source) that may be used for this application — for example, related to data residency, federal compliance, or procurement rules such as Build America, Buy America (BABA)?
Answer:
The Challenge does not prescribe or prohibit specific AI/ML model providers or require that applicants use either open-source or proprietary/closed-source models. Applicants may propose the AI/ML tools, models, software, and licensing approach that best support their solution, provided the proposal clearly identifies any proprietary limitations, licensing restrictions, third-party dependencies, data access constraints, or restrictions on DDOT’s, US Ignite’s, or academic partners’ ability to use, analyze, retain, share, or publish pilot data and results.
Applicants must also describe how their proposed approach will comply with applicable privacy, security, data governance, and procurement requirements, including data handling, storage, access controls, retention, and protection of potentially identifiable information. The Challenge pilot does not impose a specific Build America, Buy America requirement on AI/ML models or software licensing; however, applicants should identify any compliance considerations, restricted technologies, cloud/data residency assumptions, or sourcing limitations that could affect pilot performance or future procurement use.
Q4.10 Given the importance of ensuring compatibility with the analytical capabilities, are contacts/discussions allowed with the academic recipients of the data, GWU, and the University of Washington, prior to the submission date? Are they allowed to team with proposers as subcontractors/consultants?
Answer:
To ensure a fair and consistent Challenge process, applicants may not engage in proposal-specific discussions with GW or the University of Washington before submitting their proposals.
GW and the University of Washington are serving as academic partners to DDOT for the AVO Challenge and are supporting data management, analysis, and reporting for the pilot. Because of this role, they are not eligible to participate as proposers, subcontractors, consultants, or team members on applicant teams.
5. Eligibility & Participation
Q5.1 Are there any STEM/STEAM school students who would be able to participate in this program?
Answer:
DDOT welcomes opportunities to engage STEM and STEAM students, particularly through partnerships with schools and educational programs.
While the challenge itself is designed for organizations capable of deploying and operating the proposed solution, DDOT is interested in exploring educational engagement opportunities and may be able to facilitate connections with local STEM programs.
Q5.2 The package asks applicants to consider the Build America, Buy America Act requirements for future procurement contracts if scaled. For this pilot, should applicants identify BABA considerations only at a planning level, or are there immediate sourcing requirements for pilot equipment?
BABA is not required for the Challenge Pilot because the District will not be purchasing any equipment through the project. BABA considerations should be included in the proposal for planning purposes beyond the pilot period.
Q5.3 The Participant Package states that “Challenge Applicants may be based anywhere in the United States,” which suggests a US-based requirement.
Are foreign entities strictly excluded from acting as the Lead Applicant?
If so, could a foreign entity still participate as a team member, subcontractor, or technology partner under a US-based Lead Applicant?
Answer:
Foreign entities are not categorically excluded from serving as the Lead Applicant. To be eligible for contract award, the Lead Applicant must be a legally formed entity in good standing in its jurisdiction of organization, have legal authority to enter into a contract with US Ignite, maintain an active SAM.gov registration before award, and satisfy all applicable eligibility, compliance, and contracting requirements.
Foreign entities may also participate as team members, subcontractors, technology partners, equipment suppliers, or other partners under a Lead Applicant. The proposal must clearly identify each known team member and describe their role. The Lead Applicant will remain the contracting party with US Ignite and will be responsible for contract negotiation, project performance, compliance with applicable requirements, and delivery of all pilot deliverables.
Applicants based outside the United States are encouraged to identify any U.S.-based staff, subcontractors, or partners who will support deployment, operations, coordination with DDOT and academic partners, data handling, installation, maintenance, and other pilot activities in Washington, DC.
6. Proposal Development & Timeline
Q6.1 Do we need to have a built solution ready for submission on the 16th, or will we have time to build it after?
Answer:
A fully deployed solution is not required by the proposal submission deadline.
Applicants should demonstrate a credible path toward deployment during the pilot period. The current schedule allows time between the award announcement and pilot launch for contracting, development, integration, and deployment activities.
Solutions that can be deployed earlier may have opportunities for additional visibility, including demonstrations during the Transportation Research Board (TRB) Annual Meeting in Washington, DC, in January 2027.
Q6.1 Based on our cost modeling, hardware and compute/cloud GPU processing costs for the two required intersections over a 6-month period will exceed the full $50,000 award before accounting for installation, maintenance, or staffing. Does DDOT/US Ignite anticipate that the award is intended to fully fund hardware and compute costs, or is there an expectation that selected teams will supplement the award with vendor-contributed, in-kind, or self-funded resources?
Answer:
The $50,000 contract award is the maximum funding available from US Ignite for the AVO Challenge. Applicants are responsible for proposing a scope of work that can be completed within that award amount, including any necessary hardware, compute/cloud processing, operations, maintenance, staffing, and other pilot costs. Proposed solutions may be scaled or limited to reflect the available budget, but the proposal should clearly describe what will be delivered, any assumptions or limitations, and how the applicant will meet the Challenge objectives within the proposed scope.
Applicants should not assume that DDOT or US Ignite will provide additional funding beyond the $50,000 award or facilitate outside sponsorship, cost-share, or in-kind support. Applicants may identify partner-provided, vendor-contributed, or other resources in their proposal, but those resources must be secured by the applicant and reflected in the proposed pilot scope. DDOT will provide equipment and personnel support where vendor-proposed equipment needs to be mounted to DDOT infrastructure, such as approved signal poles or traffic cabinets. Real-time data delivery is not required; delayed or batch delivery is acceptable if the proposal clearly describes the delivery cadence, latency, and data quality approach.