HR001123S0011.pdf

DISTRIBUTION STATEMENT A. Approved for public release. Distribution is unlimited.

Broad Agency Announcement DARPA Triage Challenge

BIOLOGICAL TECHNOLOGIES OFFICE

HR001123S0011

November 17, 2022

HR001123S0011, DARPA Triage Challenge

TABLE OF CONTENTS

PART I: OVERVIEW INFORMATION

PART II: FULL TEXT OF ANNOUNCEMENT

1 Funding Opportunity Description

1.1 Challenge Overview

1.2 Challenge Tracks

1.3 DARPA Triage Challenge BAA Scope

1.4 Infrastructure

2 Award Information

2.1 General Award Information

2.2 Fundamental Research

3 Eligibility Information

3.1 Eligible Applicants

3.2 Organizational Conflicts of Interest

3.3 Cost Sharing/Matching

4 Application and Submission Information

4.1 Address to Request Application Package

4.2 Contact and Form of Application Submission

4.3 Funding Restrictions

4.4 Other Submission Information

5 Application Review Information

5.1 Evaluation Criteria

5.2 Review of Proposals

6 Award Administration Information

6.1 Submission Status Notifications

6.2 Administrative and National Policy Requirements

6.3 Reporting

6.4 Electronic Systems

7 Agency Contacts 8 Other Information

8.1 University Funding

8.2 Information Day

9 APPENDIX 1 – Volume II checklist

PART I: OVERVIEW INFORMATION

Federal Agency Name – Defense Advanced Research Projects Agency (DARPA), Biological Technologies Office (BTO) Funding Opportunity Title – DARPA Triage Challenge (this BAA is soliciting abstracts and proposals for Tracks A and D only, other Tracks are referenced for informational purposes – see Challenge Website for official details)

Announcement Type – Initial Announcement Funding Opportunity Number – HR001123S0011 North American Industry Classification System (NAICS) – 541714 Catalog of Federal Domestic Assistance Numbers (CFDA) – 12.910 Research and

Technology Development Dates o Posting Date: November 17, 2022 o Proposal Abstract Due Date and Time: December 20, 2022, 4:00 PM ET o Full Proposal Due Date and Time: February 13, 2022, 4:00 PM ET o BAA Closing Date: February 13, 2022 o Information Day: November 29, 2022

Concise description of the funding opportunity – The goal of the DARPA Triage Challenge is to galvanize discovery of novel traumatic injury signatures to enhance triage decision-making in mass-casualty settings. DARPA is soliciting innovative proposals in the domain of signature identification for primary and secondary triage.

Table 1. Anticipated individual awards Technical Area # of Awards Approximate Award per Team (Base + all phases)

Track A - Primary Triage Real-World Competition

Multiple $2.25M

Track D - Secondary Triage Competition

Multiple $1.5M

Prizes: Approximately $7 M Types of instruments that may be awarded – Procurement contract, cooperative agreement, or other transaction.

Agency contact

The BAA Coordinator for this effort may be reached at:

TriageChallenge@darpa.mil

DARPA/BTO

ATTN: HR001123S0011

675 North Randolph Street Arlington, VA 22203-2114

Challenge Website: https://triagechallenge.darpa.mil https://triagechallenge.darpa.mil/ mailto:TriageChallenge@darpa.mil https://triagechallenge.darpa.mil/

PART II: FULL TEXT OF ANNOUNCEMENT

1 Funding Opportunity Description This publication constitutes a Broad Agency Announcement (BAA) as contemplated in Federal Acquisition Regulation (FAR) 6.102(d)(2) and 35.016 and 2 CFR § 200.203. Any resultant award negotiations will follow all pertinent laws and regulations, and any negotiations and/or awards for procurement contracts will use procedures under FAR 15.4, Contract Pricing, as specified in the

BAA.

1.1 CHALLENGE OVERVIEW

Under the authority of 10 U.S.C. §4025 to stimulate innovations using prize competition, the DARPA Triage Challenge will use a series of challenge events to drive breakthrough innovations in the identification of physiological features (“signatures”) of injury. These new signatures will help medical responders perform scalable, timely, and accurate triage. Of particular interest are mass casualty incidents (MCIs), in both civilian and military settings, when medical resources are limited relative to the need.

The DARPA Triage Challenge’s long-term vision is 1) an initial, or primary stage of MCI triage supported by sensors on stand-off platforms, such as uncrewed aircraft vehicles (UAVs) or uncrewed ground vehicles (UGVs), and algorithms that analyze sensor data in real-time to identify casualties for urgent hands-on evaluation by medical personnel; followed by 2) a secondary stage, after the most urgent casualties have been treated, supported by non-invasive sensors placed on casualties and algorithms that analyze sensor data in real-time to predict the need for life-saving interventions (LSIs) by medical personnel. Injury information provided by these sensors could be integrated with other information about the scene to accumulate evidence about the injury mechanism and characteristics in order to enhance overall situational awareness, and to focus further physiological interventions.

To advance progress towards this vision, the DARPA Triage Challenge aims to bring together multi-disciplinary teams and industries that will identify physiological signatures and develop sensor and algorithm strategies for complex MCI settings. Teams participating in the DARPA Triage Challenge will be tasked with developing and demonstrating strategies for capturing high-value signatures for either primary or secondary triage, or for both. While aspects of the DARPA Triage Challenge involve sensors and sensor-delivery platforms, the priority is the development of physiological signatures and models to detect them, not the development of new sensor or platform technology. Signatures are defined as patterns in the sensor data that reflect or predict injuries of high importance for triage assessments.

1.1.1 Motivation

Traumatic injuries incurred in diverse military and civilian settings such as combat, attacks on gatherings, accidents, and natural disasters may be fatal if not quickly identified and treated.

Medical responders use triage protocols to rapidly prioritize casualties for immediate LSIs, such as treatment of significant bleeding and management of airway injuries. Triage is difficult under any circumstance, but MCIs in complex settings pose significant challenges to current triage capabilities. For example, overwhelmingly large numbers of casualties relative to medical resources on-hand; dangers to first-responders on the scene, such as unstable infrastructure, contaminants, or hostile fire; and environmental factors that make casualties harder to find and assess, such as rubble or darkness, may prolong the response. Medical personnel may miss the time window for implementing LSIs.

Further complicating triage in such events is a poor scientific understanding of early physiological response to traumatic injury. Physiological signatures that indicate the need for LSIs in patients who do not exhibit obvious signs of serious injury are not well established, impairing decisions about allocation of medical resources at the scene and the types of facilities where casualties should be sent for treatment.

In summary, there is an immediate need for new triage tools to support medical decision-making that can scale to MCIs. In response, the DARPA Triage Challenge seeks to enable medical responders to save lives in MCIs by 1) exploiting stand-off sensors to rapidly and autonomously provide information needed in primary triage, and 2) providing continued monitoring to predict the need for LSIs in secondary triage. The DARPA Triage Challenge will focus on life-threatening conditions that medics are trained to treat, including hemorrhage and airway injuries.

1.1.2 Challenge Goals

The primary goal of the DARPA Triage Challenge is to identify physiological signatures of traumatic injuries that can be captured with stand-off and contact-based sensors to enable medical responders to maximize lives saved and optimize the use of medical resources in complex MCI settings. The priority is development of these physiological signature-based models, not development of new sensor or sensor platform technology. Of particular interest are signatures derived from multi-modal sensing approaches which can enhance robustness against degradation of individual sensors, increase confidence through corroborating data from multiple sensors, and provide deeper insight into physiological status with complementary sensor inputs.

An additional goal of the DARPA Triage Challenge is to foster new cross-disciplinary collaborations around MCI preparedness. The DARPA Triage Challenge will bring together communities with expertise in triage and emergency medical response, tactical combat casualty care, trauma physiology, and a diverse array of sensor and algorithm technologies that will live beyond the DARPA Triage Challenge and continue to deliver innovative technologies that support medical responders. To encourage broad participation across disciplines, the DARPA Triage Challenge includes both real-world and virtual competitions. Tracks are available for both DARPA-funded and self-funded teams (section 1.2). The purpose of this BAA is to solicit proposals for DARPA-funded teams only. Self-funded teams will find information on the structure and timeline for the challenge here but should refer to the DARPA Triage Challenge website for official details and how to register as participants.

1.1.3 Challenge Description

1.1.3.1 Triage Signatures

The notional DARPA Triage Challenge primary triage setting is the first few minutes of an MCI, with automated capture and processing of data from stand-off sensors to support medical responders in evaluating an overwhelming number of casualties. DARPA expects to provide the training data and test data necessary for the workshops and competition events, though competitors https://triagechallenge.darpa.mil/ may use or develop their own training data (see 1.4.2; DARPA will not fund competitors to conduct animal or human subjects research). Sensor-algorithm systems should assess aspects of victim status that medical personnel would evaluate in this initial, rapid stage of triage, such as ability to move, severe hemorrhage, respiratory distress, and alertness. Examples of stand-off strategies could include computer vision, motion tracking, remote photoplethysmography, lidar assessment of respiratory or cardiac activity, and speech recognition, among other potential strategies.

Algorithms should be trained to integrate data from multiple stand-off sensor streams and generate a real-time assessment of victim status, most importantly the need for immediate LSIs that medical personnel are trained and equipped to provide in military and civilian pre-hospital settings, such as external hemorrhage control and airway interventions.

The notional DARPA Triage Challenge secondary triage setting is a casualty collection point, Role 1 medical facility, or emergency department, after primary triage and immediate LSIs; patient status is further refined and continuously monitored to identify changes in patient condition requiring urgent intervention. Algorithms should assess signatures indicating need for LSIs in data streams from various clinical sensors, such as continuous high-frequency vital sign waveforms from multi-modal monitors and other devices currently used in pre-hospital and emergency department assessments. Algorithms should be trained to integrate data from multiple sensor streams and generate real-time predictions of LSIs that medical personnel are trained and equipped to provide in military and civilian pre-hospital and hospital settings. Algorithm strategies could include compensatory reserve and other physiology-based concepts, data-driven models, or other approaches that could enhance secondary triage decision-making.

The DARPA Triage Challenge Primary triage domain includes real-world and virtual competitions, while the Secondary triage domain is virtual only. Additionally, a Challenge Infrastructure component includes Independent Verification and Validation (IV&V) teams and DARPA’s Research Infrastructure for Trauma with Medical Observations (RITMO) program, which provides data for use in the Secondary triage competition.

The DARPA Triage Challenge is a three-year effort with three sequential 12-month phases for Primary triage and Secondary triage in parallel, each culminating in a challenge event (Figure 1).

Full proposals submitted in response to HR001123S0011 must address all phases.

Figure 1: DARPA Triage Challenge structure and timeline. Please see Section 1.2 for full description of the tracks.

Phases 1 and 2 will end with Challenge 1 and Challenge 2, respectively, and Phase 3 will conclude with the Final Challenge. Prizes will be awarded to the top-performing teams in each challenge event based on eligibility (see Section 1.1.4). In each phase, competitors will develop algorithms that detect and characterize signatures for primary and/or secondary triage. Competitors may participate in the Primary Triage competition, or the Secondary Triage competition, or both competitions (with restrictions noted in Section 1.3 below).

1.1.4 Prizes

Table 2: Prize structure*.

Domain Tracks Challenge 1: Challenge 2: Final Challenge:

Primary Triage Track A DARPA-Funded (solicited by this

BAA)

(not eligible) (not eligible) Up to $2.5M prize pool

Track B Self-Funded

Up to $200k prize pool

Up to $500k prize pool

Track C Self-Funded

Up to $100k prize pool

Up to $500k prize pool

Up to $1M prize pool

Track D DARPA-Funded (solicited by this

BAA)

(not eligible) (not eligible)Secondary Triage

Track E Self-Funded

Up to $200k prize pool

Up to $500k prize pool

Up to $1.5M prize pool

*The dollar amount listed for a prize pool is divided among the top-performing teams in that challenge event. The number of prizes per event and distribution of the pool across top-performing teams will be determined closer to the event.

While all teams are eligible for prizes in the Final Challenge, only self-funded teams are eligible for prizes in Challenges 1 and 2. Prize pools will be divided among the top three eligible teams in the relevant competitions. The Government's obligation for prizes under the DARPA Triage Challenge is subject to the availability of appropriated funds from which payment for prize purposes can be made. No legal liability on the part of the Government for any payment of prizes may arise unless appropriated funds are available to DARPA for such purposes.

Please see Section 1.2 for full description of the tracks.

1.1.5 DARPA Triage Challenge Events

Multiple events will be held each year, with a mix of in-person, virtual, and hybrid formats (Table 3). The format is subject to change depending on public safety conditions (e.g., COVID-19 rates).

Table 3: DARPA Triage Challenge list of events.

Event

Primary Triage -

Real-World Track A

Primary Triage -

Real-World Track B

Primary Triage - Virtual Track C

Secondary Triage

Track D

Secondary Triage

Track E

Estimated duration

Phase 1 Program Initiation In-person N/A N/A In-Person N/A 2 days Challenge Kickoff Virtual Virtual Virtual Virtual Virtual 2 days Workshop - Month 8 In-person In-person Hybrid In-person Hybrid 4 days Challenge 1 - Month 12 In-person In-person Virtual Virtual Virtual 5 days

Phase 2 Workshop - Month 4 In-person In-person Hybrid In-person Hybrid 4 days

Challenge 2 - Month 12 In-person In-person Virtual Virtual Virtual 5 days

Phase 3 Workshop - Month 4 In-person In-person Virtual In-person Virtual 4 days Final Challenge - Month 12

- Preliminary Rounds In-person In-person Virtual Virtual Virtual 4 days Final Challenge - Month 12

- Finalists only In-person In-person In-person In-person In-person 1 day

1.1.5.1 Program Initiation Meeting

DARPA will hold a Program Initiation Meeting for contracted Track A and Track D teams to discuss technical approaches and execution plans, anticipated to occur in early Fall 2023 upon contract award.

1.1.5.2 Challenge Kickoff

DARPA will hold a public event shortly after the Program Initiation Meeting to mark the official kickoff of the DARPA Triage Challenge and announce the structure, rules, important dates, and prizes for the three challenge events, for all DARPA Triage Challenge tracks. Further details of the Challenge Kickoff event will be provided at a later date.

1.1.5.3 Technical Workshops

DARPA encourages vibrant information exchange and collaborative interactions among all DARPA Triage Challenge participants, to include DARPA technical staff, representatives from Challenge Teams, infrastructure developers and other Government Partners. To that end, DARPA will host at least one workshop in each phase to offer a forum for community building and cross-pollination of technical ideas and approaches:

Primary Triage Competition: In each phase (8 months into Phase 1, 4 months into Phases 2 and 3) DARPA will host a hybrid in-person and virtual workshop, with the physical site (Continental United States) identified no later than 3 months before the workshop (Track A competitors should plan to attend in-person). This will include live system integration practice sessions for real-world competitors to test their sensors and systems in simulated casualty scenes similar to the next challenge event, and virtual practice sessions for Primary Triage-Virtual competitors to test their algorithms in simulated casualty scenes similar to the next challenge event. The practice sessions will be followed by a “lessons learned” discussion for all tracks and an opportunity to discuss real-world needs with Government partners in a hybrid format.

Secondary Triage Competition: In each phase (8 months into Phase 1, 4 months into Phases 2 and Phase 3, DARPA will host a hybrid in-person and virtual workshop, with the physical site (continental United States) identified no later than 6 months before the workshop (Track D competitors should plan to attend in-person). This will include opportunities to interact with RITMO performers providing data for the Secondary triage challenge, to develop a better understanding of data-collection setting and procedures and provide feedback on the potential utility of the data for the DARPA Triage Challenge.

1.1.5.4 Challenge Events

The DARPA Triage Challenge will host Challenge 1, Challenge 2 and the Final Challenge event for both Primary and Secondary triage competitions, which will assess each competing team’s approach.

Primary Triage - Real-World Competition: Teams in Track A and Track B will bring their systems to a physical site (continental United States), to be identified no later than 6 months before each event. Track A competitors should plan (in their proposals) to travel to Challenge 1, Challenge 2, and the Final Challenge in addition to the 3 pre-challenge workshops described above.

Primary Triage - Virtual Competition, Secondary Triage: virtual track teams in Tracks C, D, and E will compete remotely via online submissions for each event. Details on submitting software and competition execution will be provided at a later date; teams can expect to submit software products for execution and evaluation by DARPA. The final round of the Final Challenge for the virtual competitors will be co-located and concurrent with the Real-World Final Challenge.

Therefore, Track D competitors should plan in their proposals for travel to attend and support their effort at the Final Challenge venue.

Post-Challenge Hot Wash Following each Challenge, DARPA will host a hybrid ‘lessons learned’ session open to all teams that competed.

1.2 Challenge Tracks

Competitors may enter the DARPA Triage Challenge via five distinct tracks outlined in Table 4.

The Challenges are organized into three competitions (Primary Triage-Real-World, Primary Triage-Virtual, and Secondary Triage) with tracks for DARPA-funded (Tracks A and D) and self-funded competitors (Tracks B, C and E).

Table 4: Descriptions of the DARPA Triage Challenge Tracks.

Problem Domain

Competitions Description Track

Track A DARPA-funded

Real World Competition

Competitors use autonomous systems with stand-off sensors to capture physiological data. Competitors’ algorithms identify injury patterns for triage.

Track B Self-funded

Primary Triage Signature Competitions

Virtual Competition

Competitors use stand-off sensor data taken by the IV&V team. Competitors’ algorithms identify injury patterns for triage.

Track C Self-funded

Track Specifications & Requirements Teams wishing to be DARPA-funded (Tracks A and D) must submit a proposal in response to this BAA. Please see Section 1.3 for proposal scope. If submitting a proposal for both Track A and Track D, proposers must submit separate proposals. Submitters whose proposals to Track A or Track D are not selected for funding will still be eligible to participate as self-funded teams in Track B, C, or E, provided that they meet qualification criteria.

Teams planning to register for Track B, C, or E do not need to respond to this BAA and should instead plan to register on the DARPA Triage Challenge website when team registration opens.

Rolling registration will be available starting one month prior to program kick-off.

Teams in all tracks will be required to qualify prior to each challenge event (qualification criteria to be posted on the website NLT 90 days prior to the qualification deadline). The qualification process is anticipated to constitute a short abstract providing details on the technical approach.

Additionally, Track A and B competitors will be required to describe safety features and may need to demonstrate certain aspects of their systems to DARPA to verify safety and operator competency. Track A and B competitors should also be prepared to host DARPA personnel for site visits to verify competition readiness.

DARPA may fund up to seven (7) teams in Track A and up to six (6) in Track D. DARPA may adjust at its discretion the number of teams accepted to compete due to technical or funding limitations, the number of applications received, or other factors.

As described in Section 1.1.4, DARPA-funded teams are not eligible for prizes in Challenge 1 and Challenge 2 but are eligible for prizes in the Final Challenge. Nevertheless, all teams in each competition are evaluated with the same criteria and scoring metrics independent of whether they are DARPA-funded or Self-funded.

1.2.1 Primary triage competitions (Track A DARPA-funded, Track B Self-funded, and Track C Self-funded)

The objective of the Primary triage competitions is to detect and identify physiological signatures of injury derived from data captured by stand-off sensors to enable early prioritization of casualties, allowing medical care professionals to quickly focus on the most urgent casualties.

Competitors will develop algorithms that detect those signatures in real-time from stand-off sensor data to provide decision support appropriate for austere and complex pre-hospital settings. Of particular interest are signatures of acutely life-threatening conditions that medics are trained and equipped to treat during primary triage, such as hemorrhage and airway injuries.

Track D DARPA-funded

Secondary Triage Signature Competition

Virtual Competition

Using sensor data provided by RITMO via IV&V team, Competitors’ algorithms predict life-saving interventions.

Track E Self-funded

Real World (Tracks A and B) Challenge events for Real-World competitors will be physical simulations of casualty scenarios.

Although the setting and complexity of challenge events will vary over the course of the DARPA Triage Challenge, the following features are expected to be maintained across events. Each competitor will have access to the same scenario, and no two (2) teams will operate in the same location simultaneously. Competitors will have only general information on the setting beforehand—for example, that it is a battlefield scenario, or a collapsed building following an earthquake. There will be actors and manikins exhibiting varying injuries and severity of injury (subject to the limitations of what can be simulated).

Competitors will use their stand-off sensors, robotic mobility platforms (e.g., UAVs, UGVs), and algorithms to autonomously process sensor data and provide real-time casualty identification and injury assessment. No part of a competitor’s system may touch a casualty or manipulate the scene (e.g., clear rubble). Each scenario will have a time limit, with no scenario expected to have a duration greater than one hour. Scoring will be based on a combination of speed and accuracy measures as described in Section 1.2.1.1.

At the beginning of each phase, DARPA will provide additional training data on the participant portal based on multimodal recordings of physical simulations of casualty scenarios similar to the end-of-phase challenge event. To assist Primary Triage Real-World competitors in developing their strategies, DARPA will host workshops at month 8 of Phase 1 and month 4 of Phases 2 and

3. These workshops will provide an opportunity for Real-World competitors to collect data from physical simulations of scenes similar to end-of-phase events. Real-world competitors should plan for 4 or 5 days of travel to each workshop and challenge event, each to be held at various locations in the continental United States. DARPA expects to provide the training data and test data necessary for the workshops and competition events, but competitors may use or develop their own training data (see Section 1.4.2; DARPA will not fund competitors to conduct animal or human subjects research).

Virtual (Track C) For Primary triage virtual competitors, DARPA will provide the same training data provided to Primary Triage Real-World competitors, starting at DARPA Triage Challenge kickoff. End-of-phase events will evaluate competitors using test data that is similar to the provided training data, but not shared in advance of the competition events. DARPA will hold virtual workshops for Primary triage virtual competitors on the same schedule as the Real-World workshops. DARPA will provide additional training data recorded at the workshops for the virtual track.

1.2.1.1 Primary Triage: Scoring Criteria and Challenge Rules

DARPA expects to release an initial DARPA Triage Challenge rules document with scoring criteria and competition rules by the Challenge kickoff. DARPA expects to apply the following context in its formulation.

The general scenario of interest for the Primary Triage competitions is an MCI where the number of casualties and/or the environment likely would preclude a timely initial assessment of each casualty by first responders. The competitions will include evaluations of autonomous stand-off approaches that assist responders in performing those initial hands-on assessments by rapidly providing information on casualty location and injury classification. The urgency in completing the course objectives and providing near-real-time situational awareness updates will be a consistent focus of the competition.

Candidate scoring criteria include the following (not in priority order and not necessarily an exhaustive enumeration):

- Percent of casualties identified

- Accuracy in reporting casualty locations

- Accuracy of casualty injury classification

- Time to complete casualty assessments

- Identification of urgent casualties before the opportunity to address LSIs would have closed

Scoring of injury classification accuracy will focus on injuries that a medic would assess in their initial hands-on evaluation, which are largely consistent across field triage guidelines. For example, a medic would rapidly assess alertness; presence of immediately life-threatening injuries, such as massive hemorrhage and respiratory distress; and presence of a non-survivable injury pattern. Competitors may adopt a particular triage guideline to provide overall priority recommendations for each casualty (e.g., “Immediate”, “Delayed”, “Minimal”, “Deceased”), and these priorities may be included in the user interface for feedback from potential end-users (Section 1.4.2), but scoring is expected to be based on detection of injuries that would be needed to establish priority in any of the triage guidelines. DARPA expects to provide the specific injury patterns that competitors should detect along with the scoring criteria at Challenge Kickoff. The injury patterns are not expected to include injuries that are not of concern in primary triage, such as minor wounds and injuries that are not life-threatening, or accuracy of vital sign measurements beyond what would be needed to ascertain an immediate threat to life.

DARPA will tune the various competition design elements to drive innovation and investment, as described below. Proposals offering approaches that are versatile and robust to the complex environmental conditions found in real-world scenarios are of high interest to DARPA.

Competitor system requirements: Real-World Competitors will develop systems to provide real-time initial assessment of casualty status based on commercial off-the-shelf (COTS) or near-COTS stand-off collection and analysis of physiological data that can scale to MCIs. In the representative use case, human responders determine the courses of medical action, but the tools developed under this effort will provide information that could be used to support those decisions. A competitor’s system must include 4 key elements: sensor packs, sensor-delivery platforms, signature-identification algorithms and processing hardware, and a user interface. The system should be fieldable by a group of up to three responders, with additional systems or elements added to scale with incident size. Therefore, all system elements together must pack down to be carried by a single vehicle (car, sports utility vehicle, pickup truck), with individual components being easily portable by a single person.

DARPA is considering the possibility of providing access to a communications network during the competition events. Prospective teams should include a description of their networking approach and specifications in their proposals.

1.2.1.2 System Elements

Sensor Packs DARPA expects that multi-modal approaches will be required to improve signature identification and address multiple challenge elements that could degrade the usefulness of any one sensor (e.g., environmental conditions, casualty pose). Various sensor modalities and combinations will be allowed, including but not limited to LIDAR, acoustic, visual, RF, IR, UV, radar, gravity, compass/magnetic, GPS, and chemical. DARPA anticipates that COTS or near-COTS sensors will be sufficient. Sensors must be capable of detecting the desired signatures from a minimum distance of 1 meter (for safety of MCI actors and may be subject to additional limitations on UAVs). All sensors elements must be skin- and eye-safe.

Platforms In Phase 1, competitors may be permitted to teleoperate their mobile robotic platforms with up to one human operator per robotic platform. In later phases, teams will be limited to only one human operator who will be the only team member that will be permitted to communicate with the robotic platforms.

Robotic platforms, such as uncrewed aerial vehicles (UAVs) and uncrewed ground vehicles (UGVs), will carry the sensor packs and any associated computation and communications hardware. Additional computation may be run on an accessory computer at a command station.

It is expected that robotic platforms will need to operate autonomously to be effective in the MCI scenarios of interest to DARPA. Some limited human interaction with the deployed robotic systems may be permitted, but DARPA expects to limit teams to only one human operator by Phase 3 of the competition. Platforms are not permitted to use tethers for power or communications, but it is expected that battery changes may be permitted. Platforms must not produce any visible illumination other than what is legally required for UAV flight.

Given the large-scale nature and complexity of MCIs, it is anticipated that a multi-agent and/or heterogeneous approach to navigating and searching the competition course could provide advantages over approaches that are limited to a single platform. The courses will be modeled after real events such as a battlefield, an earthquake with collapsed building, an explosion, or a mass shooting incident. Such situations may result in a variety of observation and navigation hazards, including rubble, immobile vehicles, constrained passages, large drops/climbs, and other casualty obscuring obstacles. By Phase 3, competitors should expect both constrained areas with human-crawlable cross sections as well as larger open spaces that could include large debris or crowds.

It is expected that teams will be limited to a maximum of five deployed robotic systems. Given the complexity of real-world environments and the expectation of constrained passages, teams are encouraged to consider robotic platforms that do not exceed 1.5m in diameter and can be readily deployed by a small team.

For UAVs, it is anticipated that the total takeoff mass per UAV is expected to be less than 9kg.

Proposals for larger or heavier UAVs should include a rationale for the decreased portability and increased safety concerns related to larger platforms. Additional weight-based restrictions on the minimum stand-off distance of UAVs will be supplied in the DARPA Triage Challenge rules document. It is expected that altitude will be limited to 30 meters. UAVs must be National Defense Authorization Act (NDAA)-compliant.

Signature-identification algorithms and processing hardware The algorithms that process sensor data and identify physiological signatures are the primary focus of this Challenge. DARPA will compare algorithm outputs to “ground truth,” which will be based on the scenario developed by the IV&V team to support scoring (e.g., accuracy of casualty detection, injury pattern classification). Competition scenarios are time-limited, and algorithm outputs will only be scored within the duration of the competition run. Therefore, algorithms should process sensor data to identify casualties and classify injury patterns in near-real time. A workstation at a “command post” external to the course may be used to run the algorithms in place of or in addition to any on-platform computation. Further information on workstation specifications will be provided in the DARPA Triage Challenge rules document.

User interface While the main analysis may occur on a desktop, competitors must provide a user interface on a laptop or handheld device for real-time visualization of algorithm results. The intended user for the interface is a medic responding to the MCI. The interface should provide the situational awareness that a medic would need during primary triage in the early moments of an MCI, such as the locations and status of casualties who most urgently require LSIs. The device and appearance of the interface may not be part of the scoring metric, but stakeholders, potential end-users, and DARPA will consider the effectiveness of the user interface in evaluating the teams’ approaches.

1.2.1.3 Technical challenge elements

The competition courses will be designed to assess performance across various challenge elements, including degraded sensing, obscuring obstacles, terrain obstacles, dynamic obstacles, and dynamic casualties. The challenges will become increasingly difficult from Phase 1 to Phase 3.

1. Degraded Sensing: The courses are expected to include elements that range from constrained passages to large fields, lighted areas to complete darkness, and wet to dusty conditions. Sensors will need to have the dynamic range to operate reliably in these environments. Dust, fog, mist, smoke, talking, flashing light, hot spots, and gunshot and explosion sounds are within scope of this challenge element. Extreme temperatures, fire, tremors, and hazardous materials are not expected to be within scope.

2. Obscuring obstacles: Casualties may be fully visible to partially obscured to completely obscured, such as buried under a shallow layer of rubble. Sensor modalities capable of penetrating rubble may have an advantage in such situations. Casualties may also be grouped with limbs overlapping, or may be interacting with live responders.

3. Terrain Obstacles: The competition scenarios will be held in realistic environments that may include natural or human-made materials; structured or unstructured clutter; and intact or collapsed structures and debris. Robotic mobility is not the focus of this challenge and it is expected that widely available outdoor robotic platforms will be capable of navigating https://www.congress.gov/116/bills/hr6395/BILLS-116hr6395enr.pdf https://www.congress.gov/116/bills/hr6395/BILLS-116hr6395enr.pdf the competition environments. In some cases, more robust platforms may benefit from more direct navigation paths or closer approach paths to casualties.

4. Dynamic Obstacles: Live responders, ambulatory wounded, or other physical changes to the environment may be included to test the agility of the system autonomy to identify and assess casualties.

5. Dynamic Casualties: Some treatable injuries may rapidly become fatal, and delays in finding and assessing casualties may result in missing the window for effective LSI. While competitors are not expected to re-evaluate casualties for changes in status, casualties who are not evaluated within an appropriate timescale may have a change in status (for example, progression of untreated hemorrhage or airway injury).

6. Endurance Limits: It is expected that individual scenarios will run between 15-60 minutes.

Teams may be permitted to replace batteries during their run, but teams should consider the implications of returning to the original launch location and redeploying their systems.

These technical challenge elements highlight key difficulties presented to systems that operate in MCI environments. DARPA is interested in proposals (Section 1.3) that offer holistic approaches to these challenge elements.

1.2.2 Secondary triage competition (Track D DARPA Funded and Track E Self-funded) The objective of the Secondary triage competition is to identify physiological signatures of injury derived from data captured by non-invasive sensors (contact-based or stand-off) to enable anticipatory decisions and prioritization for medical evacuation and care. Performers will develop algorithms that detect signatures in these data streams to provide decision support appropriate for austere and complex pre-hospital settings. Of particular interest are early signatures indicating a need for LSIs against conditions that medics are trained and equipped to treat during secondary triage, such as hemorrhage and airway injuries.

The Secondary triage competition is virtual only, and will use DARPA provided de-identified, multi-modal physiological data from trauma patients in diverse settings and cohorts provided by the DARPA Research Infrastructure for Trauma with Medical Observations (RITMO) program. DARPA will provide access to a subset of these data for algorithm training, and evaluate competitor algorithms on holdout test data in end-of-phase challenge events. DARPA expects to provide the training data and test data necessary for Secondary triage competition workshops and competition events, but competitors may use or develop their own training data (see Section 1.4.2;

DARPA will not fund competitors to conduct animal or human subjects research). Scoring may include metrics that combine speed (time to make a prediction) and accuracy measures (Section 1.2.2.1). Secondary triage competitors may be DARPA-funded or self-funded. Challenge events will become progressively more complex and realistic from Phases 1 to 3, as described in Section 1.2.2.2.

DARPA will hold virtual workshops for Secondary triage competitors on the same schedule as the Primary Triage Real-World workshops.

DARPA may fund up to six (6) teams in Track D to participate in the competitions with a total of up to thirteen (13) teams in Tracks A and D combined (see Table 2). DARPA, at its discretion, may adjust the number of teams accepted to compete due to technical or funding limitations, the number of applications received, or other factors.

1.2.2.1 Secondary Triage: Scoring Criteria and Competition Rules The scoring criteria and competition rules are expected to be released at the DARPA Triage Challenge kickoff. DARPA expects to apply the following context in its formulation.

The general scenario of interest for the Secondary Triage Challenge is an MCI where medical responders have completed primary triage, including immediate LSIs, and now must monitor casualties in an austere setting while awaiting scarce MEDEVAC or definitive care resources. The challenge will evaluate non-invasive physiological signatures that provide early warning of the need for LSIs (within the first 24 hours of monitoring) to support anticipatory decision-making about resource allocation.

Candidate scoring criteria may include the following (not in priority order and not necessarily an exhaustive enumeration):

- Accuracy of LSI prediction (for any LSI and for more specific classifications, such as hemorrhage or airway interventions)

- Lead-time for LSI predictions

- Duration of monitoring and processing needed to generate LSI predictions

DARPA may tune the various competition design elements to drive innovation and investment, as described below. Proposals offering robust, versatile approaches that can adapt to such variable adjustments are of high interest to DARPA.

1.2.2.2 Technical Challenge Elements

The challenge events will be designed to assess performance across various challenge elements, including multiple data sources, multiple data inputs, raw data, and degraded data. The challenges will become progressively more difficult from Phases 1 to 3.

1. Multiple data sources: With varied source populations, patient demographics, types of injury, and standard clinical operating procedures. Each data set may have a different standard set of sensor readings, with different added sensors in each phase of the DARPA Triage Challenge. Approaches must demonstrate robustness across a range of datasets and sensor types.

2. Multiple data inputs: Includes static data (e.g., mechanism of injury or anatomical injury pattern); multiple simultaneous, continuous streams of high-frequency waveforms; and potentially point-of-care imaging data.

3. Raw data: Data as it comes from the sensors (i.e., without any post-sensor cleaning), with the noise, aberrant values, and dropouts that occur in clinical environments.

4. Degraded data: DARPA may inject additional challenges that can be expected in battlefield and civilian pre-hospital settings, such as noisy, aberrant values, severe degradation or total loss of a particular sensor, to test the robustness of competitor strategies to such plausible scenarios.

1.3 DARPA TRIAGE CHALLENGE BAA SCOPE

This BAA seeks proposals for DARPA-funded Primary Triage Real-World (Track A) and DARPA-funded Secondary Triage competitions (Track D) ONLY.

DARPA-funded and self-funded teams will compete together in both Primary and Secondary triage Events (except for Primary triage virtual competition, which is self-funded only). DARPA will provide information for self-funded participation prior to DARPA Triage Challenge kickoff.

Proposals should include a 12-month base period in Phase 1 and 12-month option years for the subsequent phases.

Proposers may address one or both of the DARPA-funded tracks (i.e., Tracks A and/or D). If responding to both tracks, separate proposals must be submitted for each. Proposers are responsible for ensuring their team has the requisite technical expertise, capabilities, and facilities to address all aspects of their proposed technical areas. Note: As specified in the RITMO BAA (HR001122S0043), to avoid organizational conflicts of interest (OCI) or the appearance of an OCI, performers in RITMO or entities with access to data collected for RITMO will not be permitted to participate in the Secondary triage competitions unless they can demonstrate (1) measures to firewall individuals in the RITMO performer team; and/or (2) measures to firewall RITMO data from the individuals who would participate in the Secondary triage competitions. Proposers to DARPA Triage Challenge Track D must provide these demonstrations in their response to this BAA (instructions for Track E competitors will be provided by DARPA Triage Challenge kickoff).

DARPA will evaluate the adequacy of these measures to determine if the proposed measures sufficiently mitigate potential OCI.

1.3.1 Primary Triage: Stand-off casualty assessment

Track A competitors must define the modalities that will be used for stand-off capture of physiological data; the relevant data to be captured by the proposed modality; current sensor performance characteristics, including known advantages and disadvantages; and a strategy for achieving competition metrics, including mitigating any performance risks. In order to provide a comprehensive physiological assessment, competitors must incorporate multiple sensing modalities. These modalities must be considered safe for exposure to the casualties (e.g., skin- and eye-safe). Approaches must be resistant to real-world attenuators, such as low light, occlusion, and dust. They must be resilient to artifacts such as those resulting from vibration, motion, skin pigment, or topography. While any signature that aids in triage decision-making can be investigated, of particular interest to DARPA Triage Challenge are signatures that indicate a need for immediate LSIs that medics are trained and equipped to deliver in the field, including hemorrhage and airway injury. Approaches that focus solely on stand-off capture of vital signs without further analysis to support triage decisions are not of interest to this Challenge.

https://sam.gov/opp/c8d42ae0a28a4103a9ceb1d3eae57439/view

As the primary focus of the DARPA Triage Challenge is the development of physiological signatures and analytical algorithms, not hardware, proposers are encouraged to use COTS, or near-COTS sensors, and robotic platforms. However, if a proposing team has an aforementioned component under development that could be modified within the duration of the program and used in the challenge events, it can be included in the proposed approach.

1.3.1.1 Primary Triage Phases

Phase 1 (Base, 12 months) During the 12-month Phase 1, Track A teams will develop sensor packs, algorithms, and mobile robotic platforms for use in Challenge 1; engage with the IV&V team to inform design of MCI simulations; and identify risks and develop mitigation plans for Phase 2.

Additionally, teams will complete the baseline design, development, integration, and testing of their proposed solutions. The main development effort in this phase will culminate in a Critical Design Review (CDR) in month 7, during which the performer will present as deliverables a Technical Data Package. The Technical Data Package will include (as appropriate) the following items:

o Mechanical and electrical performance specifications, including weight, dimensions, power requirements and expected runtime.

o Computer-aided design (CAD) models of system(s) with all components (hardware, actuators, electronics) o Specifications on all sensors o Other design files (e.g., wiring diagrams, electrical board schematics) o Bill of materials, quotes, lead times of all major components, build plan, and schedule o Mechanical, electrical, and data architecture design, including diagrams of architectures, interfaces, protocols, etc.

o Other information that will allow the DARPA and IV&V team to evaluate design suitability

DARPA will convene a live workshop at Month 8 for competitors to test their systems on a course similar to the first challenge event, with a debrief to provide feedback to DARPA and IV&V partners on the data made available thus far and to ask questions.

Phase 1 culminates with Challenge 1, month 12, which is expected to be an approximately five-day event with teams deploying their system(s) in the competition course. The results from the event will be used by DARPA, in part, to inform the decision to fund the option for Phase 2. It is expected that at most five (5) DARPA-funded teams will advance to Phase 2.

Phase 2 (Option, 12 months):

During the 12-month Phase 2, Track A teams will continue their development efforts and participate in the second challenge event; continue engagement with the IV&V team; and identify risks and develop mitigation plans for Phase 3. Sensor-delivery platforms under autonomous control are required in Phase 2.

DARPA will convene a live workshop at Phase 2, Month 16 for competitors to test their systems on a course with similar features to Challenge 2, with a debrief to provide feedback to DARPA and IV&V partners on the data made available thus far and to ask questions.

Challenge 2 (month 24) will be more difficult than Challenge 1, with multiple scenario types and increasing difficulty in technical challenge elements, including the need to localize causalities before identifying signatures for triage, low lighting or darkness, and partial or full occlusions.

Team performance in Challenge 2 will be used by DARPA, in part, to inform the decision to fund the option for Phase 3. It is expected that at most four (4) DARPA-funded teams will advance to Phase 3.

Phase 3 (Option, 12 months):

During the 12-month Phase 3, Track A teams will continue their development efforts and engagement with the IV&V team, and participate in the Final Challenge. DARPA will convene a live workshop at Phase 3, Month 28, for competitors to test their systems on a course with features similar to those of the Final Challenge, with a debrief to provide feedback to DARPA and IV&V partners on the data made available thus far and to ask questions.

The Final Challenge will occur in 2 rounds at Phase 3, Month 36. The Final Challenge will include a preliminary round and a final round.