Skip to content

U.S. DOT FY 2022 Phase I Pre-Solicitation Q&A


* * * 2/2/22 Update: The Pre-solicitation is now closed. Topic questions are no longer being accepted. * * *


This U.S. DOT Small Business Innovation Research (SBIR) Pre-solicitation Q&A Forum provides small businesses an opportunity to submit clarifying questions on the topics published in the U.S. DOT Fiscal Year (FY) 2022 Phase I Pre-Solicitation. Questions will be answered by the Technical Experts/Topic Authors at the U.S. DOT Operating Administrations during this period. Questions shall be limited to improving the understanding of a particular topic's requirements and should not share proposal ideas nor proprietary information.
 
The U.S. DOT’s FY 2022 Pre-Solicitation period is January 5, 2022 through February 2, 2022 at 5:00 pm Eastern Time (ET). Only questions posted within this period will be responded to.

How to Use the SBIR Pre-solicitation Q&A Forum:

Register : Anyone can view questions but registration is required if you wish to post a question and receive a notification when it is responded to.

  1. To create a new account or to sign in with an existing account please click on “sign in” located at the top right of the page .
  2. A new pop-up window will open to make a selection, you can use an existing Google or Facebook account to access the site or you might select to create a new account.
  3. If you choose to access UserVoice with an existing Google or Facebook account, you will be required to enter your login credentials for Google or Facebook and agree to terms of service
  4. If you need to create a new account, click "Create an Account" underneath the password box.
  5. In the next screen, type in the email and click "Verify Email".
  6. Navigate to your inbox and click the link within the email to verify. The link will be generated by noreply@trymagic.com.That link will be live for 20 minutes. Once that time elapses, you will have to request another one
  7. Navigate back to UserVoice, and you will need to provide your name, a password, and agree to terms of service and storage permissions. Click "Create Account".
Search: Browse the existing questions by category to find the topic you are interested in and to see if your question has already been addressed. If you see a question you are also interested in, you can add a vote to the question, or leave a comment.

Post: If your question has not already been posted, use the “Enter your idea” box to enter your question.

  1. Select the research topic your question is related to from the drop-down menu that appears when you begin typing.
  2. You may add more detail in the Description box (optional). Do NOT attach any files. The U.S. DOT shall not consider any submitted materials other than questions.
  3. Enter your email address.
  4. When finished, select “Post idea.” If you registered, you will be able to track responses to your post. Please include only one question or comment per post to help us keep the forum organized.
Please note, all comments will be public information. Potential offerors shall not include in the question(s) or comment(s) any information that they do not wish to be made public. Telephone inquiries or meeting requests will not be addressed and may be removed. The U.S. DOT shall not respond to requests for advice or guidance on any offeror’s solution to any SBIR topic and shall not consider any submitted materials other than questions.

Thanks for joining our SBIR Pre-solicitation Q&A Forum.

If you experience technical difficulties, please contact us at DOTSBIR@dot.gov.
  • Hot ideas
  • Top ideas
  • New ideas
  • My feedback

14 results found

  1. Price Elasticity for fines issued by Speed cameras

    Finding payment propensity via AI/ML could help disincentivize speeding while preventing lower-income motorists from spiraling into a debt trap. Motorists from all over the US travel our roadways and creating a system to target the appropriate fine levels based on sociodemographic and larger economic factors or trends could help the department meet its goals of safety and protect our motorists from excessive fines and fees. Local governments have struggled with fines and fees and a large scale speed camera program needs to have fiscal equity, sustainability, and fairness.

    1 vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  2. Scope of Trajectory Conflicts

    On the topic of identifying and predicting trajectory conflicts, should we restrict our focus to only conflicts resulting from vehicles caught in the dilemma zone, or are we to consider all conflicts occurring within 75 ft of the intersection stop bars? e.g. are conflicts arising from left turns within the scope of the SBIR?

    3 votes

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  3. Metrics of Success

    There are two main points of study in the pre-solicitation: identifying trajectory conflicts, and identifying when a vehicle is in the dilemma zone.

    In the pre-solicitation notice, it mentions a success metric of less than 20 percent false-positive rate for trajectory conflicts. Is there a similar metric of success when determining whether a car is in the dilemma zone? If not, should we provide a metric based upon our particular approach, and include reasoning about the application of the metric?

    2 votes

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  4. Defining Dilemma Zone

    In the pre-solicitation notice, the dilemma zone is “measured in traveling time/distance range to the intersection stop bar, within which about 90 percent of drivers will decide to brake whereas 10 percent of drivers will decide to accelerate at the time of phase change …”.

    A question had been previously asked on this forum, pertaining to the definition of dilemma zones, and the response defined the zone for each car individually, based on their individual speeds; a car is in the dilemma zone when they will reach the intersection stop bar in about 2.5s - 5.5s.

    The overall goal, as…

    2 votes

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  5. Consideration of Signal Phase and Timing Control

    The SBIR is specific for signal controlled intersections, as that is where the dilemma zone occurs. But how much, if any, consideration should be given to methods adapting signal timing based upon the results? For example, if the system predicts which vehicles will be in the dilemma zone, should it also suggest signal timing to minimize the number of vehicles that will be in the dilemma zone?

    3 votes

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  6. Camera Assumptions

    What can the offerors assume regarding the functionality, type(s) and specification(s) of the deployed cameras? (e.g., minimum and maximum framerate, resolution, and field of view).

    1 vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  7. How many topics can we apply for?

    Can companies apply for multiple research topics? For example, could my organization apply for both Topic 22-FH2 and Topic 22-FT1?

    1 vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  8. Defining the Conflict Zones

    The current videos from cameras do not depict an area for analysis (some type of marker) to identify the "Conflict Zone" under AI inspection. Is it implied that a method will be included in the implementation design to provide this information, or will that be provided in some manner by the local DOT?

    2 votes

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  9. Definition and location of an Edge Server

    The boundaries of what is being defined as an Edge Server in an edge computing environment matter to reach the desired number of intersections and videos being processed. Is the intent to bound the definition to a single physical box? And is there a requirement as to where the server should be located?

    1 vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  10. 1 vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  11. Geographic intersection limits?

    Is this project limited to one geographic location with one or more intersections or is there an option to compare it against other locations with different situational aspects relating to traffic systems and multi-modal patterns?

    1 vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  12. 20 Percent False Positive

    The draft solicitation states "If the prototype deems successful in logging and classifying dilemma-zone-related trajectory conflicts at less than 20 percent false positive rate..." It would be helpful if a definition could be provided for what constitutes a false positive. For example, is it considered a false positive if an algorithm identifies a trajectory conflict between two vehicles, but the conflict is averted because one of the vehicles turns?

    2 votes

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  13. Existing Processing Resources are Implied

    ​​Given that this research project will "...develop an edge server-based AI application," but apparently not the edge computing platform, what assumptions can the offerors make about the resources of the edge computing platform? This is critical because artificial intelligence applications for video analysis generally require specialized Graphics Processing Units (GPUs) in addition to standard Central Processing Units (CPUs) found in business processing computers. Parallel algorithms applied simultaneously to multiple video sources will also require a number of GPUs that is proportional to the number of video sources (not necessarily one-for-one, but it should be obvious that the processing capabilities of…

    1 vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

  14. Dilemma Zone detection with multiple technologies

    Radar, Lidar and Video imaging have made huge strides in the past few years - I think this topic should be open to other technologies or fusion of other technologies. The pre-selection of video, or video only appears too narrow and not in tune with recent advances in the field. I suggest keeping the theme, but broadening the allowable approach.

    3 votes

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    We’ll send you updates on this idea

Feedback and Knowledge Base