Autonomous triage and evacuation

The outcome should contribute to:

• Increase knowledge on technology and requirements and accelerate the development of life-saving technologies by capitalising on the unmanned vehicle development and develop complimentary defence medicine related autonomous functionalities.
• Proof-of-concept demonstrations highlighting the possibilities and potential limitations with autonomous vehicles for casualty transportation and triage.
• Create an R&D technology development roadmap for RAS CASEVAC platforms.
• Improve accuracy and speed in locating and evaluating casualties.
• Reduce risk exposure for combat medics and medical personnel.
• Human-machine teaming technology development.
• Dismounted soldier system development.
• Ethically acceptable decision algorithms.

General objective

Large Scale Combat Operations (LSCO) between peer adversaries can result in mass casualty scenarios where the need for casualty care and evacuation dramatically outstrips available medical resources. Unmanned air, ground and sea vehicles could significantly improve evacuation capacity and enable rapid automated or fully autonomous battlefield triage, also under Chemical, Biological, Radiological and Nuclear (CBRN) conditions and in high intensity fighting areas, resulting in faster and more efficient care, increasing life and limb saving opportunities in the early stages of the evacuation chain. This call topic therefore addresses the urgent need to develop and validate innovative Robotic and Autonomous System (RAS), i.e., autonomous and robotic-assisted capabilities that address the specific challenges of military battlefield triage and evacuation in mass casualty scenarios, including CBRN contamination and ongoing high intensity fighting spots with limited or no access of first responders.

Specific objective

The development of unmanned military platforms for surveillance, reconnaissance and kinetic attack missions is progressing rapidly. Unmanned systems have the potential to substantially increase the RAS CASEVAC evacuation capacity in mass-casualty scenarios, including CBRN contamination and ongoing high intensity fighting areas with limited or no access of first responders, and to expedite the triage, diagnostic and initial treatment process from the point of injury. The concept of autonomous triage in LSCO should be based on life threatening indicators as a minimum (i.e., covered in the START algorithm). However, the realisation of such capabilities requires development of dedicated solutions that provide innovative damage site inventory of casualties, extraction, and unmanned systems (various platforms) with the ability to monitor and assess the health status of injured soldiers and adapt their behaviour accordingly. RAS CASEVAC platforms need to be able to continuously adapt their route and speed to all environment and weather conditions, unexpected events, threat level and the condition of on-board patients, while providing physical protection.

This call topic targets two technologies that have the potential to save lives in LSCO mass-casualty scenarios, namely: (i) RAS within the CASEVAC system including autonomous battlefield triage, and (ii) autonomous CASEVAC system improving the overall logistics chain to/from the battlefield. The required novel functionalities include safe transportation of casualties to a suitable medical treatment facility following the golden hour timeline whilst providing a basic level of physical protection (towards shrapnel, small-arms fire and all-weather conditions) and patient monitoring.

Proposals must address the development of a RAS CASEVAC multi-role approach. Time is of the essence in LSCO missions so the proposals must explore easy re-configurations concepts and compatibility between different payloads – medical and non-medical – through an Interoperable Modular and Scalable Architecture (IMOSA) approach. This allows quick interchangeability of components and interoperability between different missions for the autonomous platform, including a “plug-and-play” capability for (wearable) monitoring sensors and (wearable) patient care sensors, combined with remote patient assistance.

Proposals must address RAS within the CASEVAC system including autonomous battlefield triage. Particular attention should be paid to trusted autonomy for effective networked and autonomous and automatic CASEVAC missions, including a swarm-based manned-unmanned teaming (MUM-T) in demanding denied/contested environments.

The possibility of standardised interfaces should be explored to allow the integration of a variety of patient monitoring and CBRN-sensors to be used in different configurations depending on the CASEVAC mission, and to facilitate the use in defence, civil and dual-use configurations for efficiency in the logistics chain (evacuation chain). In addition, proposals may also address the potential synergy for use by law enforcement and other governmental use.

Proposals must:

• Evaluate integration of sensors of the wounded soldier status during CASEVAC. This includes “plug-and-play” C2 to/from the chosen CASEVAC platform and integrate monitoring of patients during CASEVAC.
• Include a comprehensive model of the physiological evaluation of the casualties, which may be fed asynchronously with information acquired from the casualties health status and from the surrounding environment. Information needed to forecast the route and adapt the autonomous system behaviour in line with the degree of injuries.
• Evaluate integration of miniaturised sensors for CBRN detection and identification and monitoring (DIM).
• Define the specific autonomous platforms to be used to provide RAS CASEVAC and START capabilities.

Proposals should also:

• Reflect on different concepts of autonomous triage from an ethical perspective, but also regarding the perspective of responsibility. The concept of autonomous triage in LSCO should be based, as a minimum, on life threatening indicators covered in the START algorithm.
• Address explainability of the forecasting and of the assessments obtained through automated procedures.
• Foresee detailed alternative approaches to the assessment of casualty status, especially in view of the lack of large databases on which Artificial Intelligence (AI)-systems can be trained.
• Increase the casualty evacuation capacity and the triage process expedition at the point of injury.
• Adapt to prevalent weather and environment conditions, threat levels, and the condition of on-board patients for autonomous casualty evacuation platforms, in addition to react to unexpected events that might happen in the local environment during navigation.
• Define methods to achieve physical protection for patients and systems during evacuation.
• Define methods to achieve platforms’ survivability.
• Address platforms’ reusability (e.g., CBRN DIM and decontamination).
• Examine the potential of fully autonomous battlefield triage, based on innovative AI-based algorithms.
• Remain operational in all weather conditions, including sub-zero temperature and snow-covered conditions.

In addition, proposals may address:

• Real-time multimodal fusion of field-collected information to provide a comprehensive and accurate situational overview.
• Integration and suitable graphical tools for cooperation among different specialists, services and C2 systems, considering federated mission networking as standard for interoperability.
• Existing platforms (UxV) for surveillance, reconnaissance, and missions, as they relate to increasing evacuation capacity and expediting triage processes.
• Capability of autonomous triage and evacuation platforms’ self-defence (e.g., navigation in mined areas, C-UAS and MANPADS missiles self-defence).

Types of activities

The following types of activities are eligible for this topic:

Accordingly, the proposals must cover at least the following tasks as part of mandatory activities:

• Study:
  • The feasibility of AI-based autonomous triage following the START algorithm, as a minimum.
  • RAS CASEVAC in harsh conditions and across contested environments, where GNSS signal and communication links could be denied.
  • Means to allow swift and safe casualty extraction from the ground and “hand-over” between different types of platforms and or operators.
  • Automatic or autonomous functions to optimise platform behaviour (planning and operation) based on risk assessment and available (sensor) data considering:
    • Patient condition and stabilisation efforts.
    • Time to destination related to golden hour elapsed time.
    • Patient condition deterioration related to speed dependent transport performance (e.g., risk of opening of wounds due to platform vibration, shorter route in difficult terrain vs. longer route in easy terrain).
    • Resource management at destination.
    • Threat level.
  • Definition of the system and system of systems (swarming) architecture gathering functional and non-functional requirements for the individual systems (UAVs, UGVs, USVs) and the overarching system of systems, evaluation technologies, specifying swarming behaviours, ensuring interoperability with standards, and assessing risks.
• Design:
  • Proof-of-concept technology demonstrations and evaluations of the (separate) developed functions for health status indicators. These should be performed in representative military scenarios.
  • Autonomous triage reflecting START algorithm.
  • Monitoring during RAS CASEVAC transport, multi-modal casualty transport with physical safety measures and adaptive behaviour.
  • RAS CASEVAC.
  • Showcasing the applicability of proposed solutions in military structures and the military decision-making process, by implementing them in the EU hosted wargaming simulation/exercise (e.g., by one or more partnering or associated Ministries of Defence, HEDI).
  • Develop a proof-of-concept mission planning tool that integrates inputs from all systems to create a cohesive operational plan. Enable real-time updates to the mission plan based on incoming data and changing conditions.

In addition, the proposals should cover the following tasks:

• Study:
  • The feasibility of autonomous or robotic-assisted systems for initial stabilisation of casualties before extraction and transport, e.g., to control haemorrhage.
  • Integration of commercial wearables into CASEVAC platform related to health monitoring of patients (health ring, electronic ID-tags, RFID-tags, smart-watches, smart-textile, etc).
• Design:
  • RAS medication during transport based on the casualty monitoring data, according to the improved first aid spectrum, e.g., painkillers, CBRN medical countermeasures.
  • Potential for a more autonomous battlefield triage, using innovative AI-based algorithms, should be examined.

The proposals should substantiate synergies and complementarity with foreseen, ongoing or completed activities in the field of medical logistics, notably through EU funded actions related to mass casualties scenarios.

Functional requirements

The proposed product and technologies should meet the following functional requirements:

• Plug-and-play capabilities.
• Miniaturised sensors.
• Either stand-off (non-contact) sensors on a small UxVor Quadruped, or wearable biosensors that are already worn or autonomously (without human intervention) placed on the casualty during the triage process.
• Robust estimation of health status indicators and vital signs in realistic battlefield conditions, including day and night, on soldiers equipped with body armour and camouflage face paint and with body movements characteristic for injured soldiers.
• RAS battlefield triage (e.g., using innovative AI-based algorithms).
• RAS allocation of evacuation priority, at the point-of-injury (PoI) and casualty collection point (CCP), based on the estimated vital signs and indicators, using existing battlefield triage methods adhering to current best practices for mass-casualty triage.
• Autonomously detecting and localising casualties at PoI, in all weather and visibility conditions, using sensors on unmanned vehicles.
• Continuous monitoring of casualties during transport using body-worn wearables or stand-off sensors, including the ability to provide alerts if health status deteriorates.
• DIM of CBRN injuries including application and monitoring of indispensable antidot-therapy.
• Protecting casualties from harsh weather effects (rain, wind, extreme temperatures) and enemy fire during CASEVAC.
• Autonomous platform provided with onboard data processing to filter and preprocess data before transmission.
• Capability to real-time and low-latency communication link between the command centre and all deployed units.
• Integration of data from multiple sources to create a unified, coherent picture of the field situation.
• Ability to detect, classify, and track objects of interest (e.g., injured individuals, obstacles) with high accuracy.
• Coordinated triage actions between different platforms (swarming).
• Adherence to relevant standards and protocols to ensure interoperability with existing defence systems.