Notice

Competition document: autonomy in hazardous scene assessment

Updated 25 October 2016

1. Autonomy in hazardous scene assessment

In this Centre for Defence Enterprise (CDE) themed competition we want to investigate the use of autonomous systems to assess scenes that are potentially contaminated with hazardous materials.

This competition was briefed at the CDE Innovation Network event on 22 September 2016 and at a webinar on 3 October 2016.

Your proposal must be received by CDE by 5pm on Thursday 10 November 2016. Your proposal must be submitted to CDE online.

CDE is running this themed competition on behalf of the Ministry of Defence (MOD) and the Home Office (HO).

Up to £1.5 million is available for phase 1 of this competition. We expect to fund a number of projects around £40,000 to £80,000 lasting up to 6 months, with phase-1 deliverables completed by August 2017.

Up to £1 million will be made available for phase 2 of this competition, for projects lasting no more than 12 months. Phase-2 funding will only be available to projects funded at phase 1 and will be considered on a per-project basis. Collaboration between successfully funded projects for enabling technologies at phase 1 will be encouraged to develop systems at phase 2.

2. Background

There are significant safety risks to human responders (police and other emergency services, armed forces) who assess potentially contaminated scenes. These hazards can be particularly severe if the incident involves chemical or biological warfare agents, or large amounts of harmful industrial materials. Risks increase further in difficult environments such as in enclosed spaces, unstable structures, fire, deep water or if the scene contains active and aggressive hostile elements.

Detection, identification, monitoring and sample (DIMS) recovery capabilities are used during scene assessment to rapidly, and with as little doubt as possible, determine the nature and extent of the hazards present. This will allow the scene commander to act with best-available information, and so make the best possible decisions regarding recovery and restoration.

Providing scene assessment is a specialist task. The MOD and HO have to select, train, equip, and exercise specialist responders. The number of responders available can affect the time required to assess a scene, depending on its size, which might vary from a stadium to a single room. There could also be a number of simultaneous scenes to assess, which adds an additional burden.

Currently our DIMS capability is provided by the specialist response teams from the emergency services, and MOD teams that provide homeland support. For this competition we’ll refer to all these teams as specialist responders, to differentiate them from the emergency service personnel who usually arrive at the scene first. Those first on the scene, who we’ll refer as initial responders, may not be carrying, wearing or trained to operate equipment that would allow them to manage a hazardous scene.

All responders are under time pressure and some may be wearing protective equipment that can considerably increase physical and mental stress, making them less able to complete their task efficiently - particularly on extended or demanding tasks.

We’re looking for technologies that remove the need to have a person in a potentially hazardous environment, while still assessing the scene as effectively. This will free the responder to conduct other activities, while also keeping them safe.

We also want to improve the amount and quality of information available to all levels of a response team (such as initial and specialist responders, the scene commander) as quickly and as close to the start of the incident as possible. For example those first on the scene need to know if there’s a hazard, where to set up the cordon, and if there are any casualties. The system should provide this as well as more detailed information, to the specialist responders before they get to the scene, so they’re prepared to act immediately on arrival.

Using autonomous systems could mean a much lower risk to life, while maintaining, or even improving, the mission outcomes when assessing high-hazard environments. The robotic platforms used in autonomous systems may be hardened to resist the effects of hazardous materials. They may also be able to perform tasks consistently and efficiently over long periods, as they don’t need the same protective measures or suffer physical or mental stresses in the same way as a human operator.

Autonomous systems could also assist specialists by removing the need for their constant attention – potentially even on complex tasks. They may well offer a better solution than traditional remote methods (using robotics alone) as these still rely on highly-trained operators, and performance could be prone to operator fatigue.

3. The challenge

This competition is made up of 2 phases. In phase 1 we’re looking for proposals for innovative technologies and systems that can effectively contribute to, and address, the tasks set out in the challenge. During phase 2, we’ll be looking to demonstrate the system solutions to show how effective they are at achieving the challenge tasks in test environments of increasing complexity.

We want future autonomous systems to be able to work together effectively with responders in human-machine teams to indicate the presence of any hazardous materials, confirm the nature of those hazards, map where those hazards are, monitor changes in both hazard and environment, and take samples of material for further analysis.

This should all be done without the need for an expertly trained operator, be affordable to buy and easy to maintain throughout life. Therefore, in this CDE themed competition we’re looking for research proposals for enabling technologies and systems that could make this possible.

This is really challenging and falls into 4 broad tasks: survey, refine, map, and sample, each of which are required to create a full capability. We don’t expect you to be able to provide a complete scene assessment system at this stage, your proposal must clearly identify which task, or tasks, your proposed solution aims to meet and to what level.

3.1 Survey

We want autonomous systems to conduct a rapid initial survey of the scene, to assist the commander in deciding what immediate steps should be taken, to help set an effective cordon, to supply information to hazard modellers, and to the specialist responder before they arrive at the scene - so we also want autonomous systems that:

• provide real-time audio-visual feedback
• indicate whether a hazardous material might be present in the air
• sense hazards typically encountered in confined spaces (such as low oxygen, flammable gases)
• provide meteorological information (especially wind direction, wind speed)

3.2 Refine

Following the initial rapid survey we want autonomous systems to undertake a more detailed survey of the target – to identify if there are physical obstacles to follow-on DIMS capabilities, to determine if there are casualties or other personnel on-scene, to help the commander maintain an effective cordon, to improve initial hazard predictions and address any unusual responses noted during the Survey task. So we also want autonomous systems that:

• continue to detect in real time, reporting any significant changes in the hazard and the environment
• resolve any unusual responses, by using higher-fidelity sensors and/or outputs from a network of complementary sensors
• create a digital representation of the scene, overlaying onto existing imagery or map data if available
• provide more detailed meteorological information (including temperature and relative humidity)

3.3 Map

Whilst the initial survey is being refined, we also want autonomous systems to map safe routes of approach to the scene by foot, and to determine which surfaces are potentially contaminated with hazardous material. We want autonomous systems to:

• examine exposed surfaces within the scene, to determine whether a hazardous material might be present on them
• establish and then mark the extent of any contamination boundaries, and ideally establish the distribution of hazard within those boundaries

It’s not always possible to determine whether a surface is contaminated by sensing for materials in the air. Sometimes methods that make contact with the surface, or use examine the surface without contact, are needed. Whichever approach is used, mapping should disturb the scene as little as possible, it shouldn’t cause clean areas to become unwittingly hazardous or lead to unnecessarily spread contamination from one location to another within the scene, or transfer the hazard outside of the already-contaminated area.

3.4 Sample

Finally, we want autonomous systems to collect, package and recover samples from the scene – possibly as part of the evidence collected in support of an investigation, possibly to improve our confidence by confirming our earlier conclusions by further analysis, and possibly to allow detection of materials that are otherwise not easy to sense. So we want autonomous systems to:

• safely collect, package and recover air samples - using sorbent media, particulate traps, or similar approaches
• safely collect, package and recover samples from surfaces using contact sampling methods

In all cases the recovered samples should be tagged with position and time of recovery. The system must be able to collect, package and recover samples from locations specified by the user (operator/commander), and also be able to collect, package and recover multiple discrete (not aggregated) samples across an area.

Samples should be recovered in a container that encloses the sample so that it does not leak, and so that the container’s outer surfaces can be decontaminated without affecting the contents.

Collecting a sample should also disturb the scene as little as possible, it shouldn’t cause clean areas to become unwittingly hazardous or lead to unnecessarily spread contamination from one location to another within the scene, or transfer the hazard outside the already contaminated area, unless deliberately intended and properly managed.

3.5 Considerations

In conducting all of the tasks, you must bear in mind that the Survey task should begin as soon after the incident is believed to involve hazardous materials as possible, ideally within 5 minutes.

The Refine, Map and Sample tasks should be under way (if required) ideally well under 30 minutes after specialist responders arrive on scene.

All tasks should be performed irrespective of whether maps, plans or pictures are already available for the scene.

All tasks should be performed irrespective of weather conditions, day or night – noting that there is usually degradation in manned capability during inclement weather, or at night.

If unmanned air systems (UAS) form part of your proposal you must comply with existing regulations. Whilst the legal and policy areas around UAS are rapidly evolving, you must work with existing regulations to ensure that, should a system develop, it is not later ruled out due to non-compliance.

The operator or commander should always have the option to intervene to gracefully bring the mission to an early conclusion; either by leaving the elements of the system on-scene or by recovering them safely.

4. What we want

In this CDE themed competition we’re looking for innovative technologies that will allow the development of autonomous systems for performing the DIMS-related tasks listed above.

Our focus is on autonomous, rather than remotely-piloted or operated systems that simply provide separation of the operator from the hazard.

We want a future system to be operated by a trained but non-expert person: in other words it doesn’t need to be operated by a dedicated expert.

Proposals can be based on an existing technology but must clearly show an innovative adaptation to fit the requirements of this competition. We’re not looking for demonstrations of existing platforms or development of new platforms, but proposals should include integrating sensors onto autonomous platforms.

Your solutions should use open data and communications architectures whenever possible, as a future system shouldn’t require bespoke networks. Ideally your system should use common interfaces.

Your system should maintain communications and continue operating whether indoors or outdoors, whatever the weather, including area with no GPS signal. In terms of size, you should assume that a platform might typically be deployed from a motorbike or backpack (man-portable) through to an estate car, commercial van, even to a small ISO container. The important point is that the platform should be appropriate to the task.

We’re particularly interested in innovative research proposals that help autonomous systems to meet the following needs:

• provide information from the scene to the responders as early as possible, without adding to the training or equipment burden of the initial responders: giving the commander, initial responders and the subsequent specialist response an enhanced awareness of the immediate situation, delivered through existing command, control and communications channels, without any burden of interpretation

• reduce the risk of sustaining casualties by reducing the number of manned human approaches required, by reducing the need for specialists to be in close proximity or direct contact with the hazard; and ideally to completely remove specialist from the hazardous area under all circumstances

• reduce our current level of reliance on the availability of specifically selected, trained and equipped individuals, ideally replacing them in the most hazardous of roles

• enable effective performance of tasks that would otherwise place excessive demands on resource, in terms of numbers of trained personnel required or their duration of employment – examples include large, complex scenes; multiple concurrent scenes; and long term wide area monitoring

• provide a consistent, predictable outcome to tasks that might otherwise be prone to physical or psychological burden, especially on extended duration tasks

• improve the speed with which the commander’s critical information requirements are met, by freeing specialists to undertake other tasks, leading to an overall reduction in time on task and swift provision of information to the commander

• reduce and ideally remove the need for human-in-the-loop processing of sensor data in generating information that can be acted-upon by the commander - by ensuring data from sensors is relayed across appropriate communications networks to most effective processing capabilities, and that the resulting information is added to a shared information picture made available to all relevant levels of command – ideally without need for human intervention.

5. What we don’t want

In this CDE themed competition we’re not looking for the development of new sensors, but proposals could include optimising sensors for use on autonomous systems.

We’re also not looking for the development of new sensors, but proposals could include optimising sensors for use on autonomous systems.

We’re not interested in any proposal that relies on a human operator being ‘at risk’ (in the hazard). This means that assistive robotics (such as load-bearing exoskeletons) or wearable technologies (such as physiological monitors) are out of scope, if they’re meant to be worn by a human operator in the hazardous area.

This CDE themed competition is not looking for the development of ‘swarming’ autonomous systems, but intelligent teaming between small numbers of platforms is in scope.

We appreciate that several of the techniques developed here could be valuable during the restoration and recovery phases. However, hazard mitigation or remediation of the environment, is out of scope for this competition.

Purely stand-off (typically optical) and through-barrier technologies are out of scope. This competition is focussed on a scenario where a hazardous material has already been, or is being, released, rather than pre-event, still-contained hazards.

We’re not looking for high maintenance solutions (requiring frequent and/or expensive servicing) or those with disposal issues, such as radioactive sources.

Finally your proposal shouldn’t need utilities, consumables or logistics support outside those typically available to the emergency responder. This might include small portable single-phase 240V AC generators, but would not include for instance liquid nitrogen coolant.

6. Exploitation and benefits

Through this competition we want to trial different approaches in relevant environments. This competition will demonstrate the benefits and advantages of using autonomous systems in assessing a hazardous scene to MOD and HO stakeholders. This competition isn’t yet tied to a specific procurement route; however it’s our aim that it will build a strong case for the systems solutions developed through the competition. The clear benefits of relevance to MOD and HO are:

• better understanding of what is practically attainable, given current state-of-the-art
• insight into what may be possible, given modifications to the systems demonstrated and/or given likely developments in the field
• understanding those approaches that work well, those that need improvement, and those that are flawed

The science and technology evidence generated through this competition will be used to underpin any future requirements involving autonomous systems for scene assessment (HO) or specialised CBRN roles (MOD).

We’re already making plans for a third phase of the competition aimed at demonstrating how concepts successful at phase 2 could be further developed and transitioned into a DIMS capability, through delivery of prototypes to appropriate demonstration unit (MOD and HO) for end-user trials.

We’re seeking research proposals that address our technical challenges and for successful phase-1 projects to be demonstrated against a relevant but straightforward set of challenges which should be clearly described by the bidders in their proposal. This will indicate their potential for integration into a more capable system, able to be demonstrated against more complex challenges at the end of phase 2.

We’re really keen to encourage collaboration between component and subsystem suppliers for phase-1 proposals, and we’ll also encourage successful bidders at phase 1 to team with fellow applicants to deliver a phase 2. Bidders may form consortia for phase 2, or blend with solutions from a subcontractor. However, bear in mind that we won’t fund the same piece of work from the same company twice.

A LinkedIn group has been set up to help potential bidders collaborate.

All projects funded at phase 1 will be expected to take part in a 2 networking events and a presentation day together with the other funded organisations. These events will provide an invaluable opportunity to meet other participants and discuss potential collaboration. You should cost attendance at these events into your proposal.

The first event will take place in February/March 2017 at or near a Dstl site where you will have the opportunity to meet the other project teams that were successfully funded at phase 1 and start to develop a collaborative community. You’ll give a short presentation (just a few slides) describing your project, what you hope to achieve during phase 1 and outline what you think your phase-2 work will look like. In order to integrate your technology with other solutions to contribute to a final system we’re really keen to promote collaboration.

The second event will also take place at or near a Dstl site in May/June 2017. We aim to provide even more opportunity for you to collaborate by introducing you to other organisations that could provide support to projects in developing their phase-2 proposals. Dstl will identify these organisations and make all phase-1 winners aware of the benefits they can offer.

The final event will take place in June 2017 when bidders will demonstrate the successful outputs of phase-1 projects to stakeholders.

As a deliverable of the phase-1 project, successful bidders will be expected to produce a fully costed proposal for a phase-2 project which must be submitted using the CDE online submission service by 5pm on 8 August 2017.

We aim to take forward a number of the most successful outputs from phase-1 projects for phase-2 funding. Only bidders funded at phase 1 qualify for entry into phase 2 of this competition where up to an additional £1 million of funding will be made available. Phase-2 funding will be awarded on a per-project basis. Phase-2 projects can last up to 12 months.

In phase 2 we’re exploring the system challenge and as we’ve mentioned, we expect to take forward a number of the most successful phase-1 projects through to phase 2. You should plan for your phase-2 project to run for up to a year and finish with final systems demonstrations in test environments of increasing complexity.

At the end of phase 2 you’ll need to demonstrate that the key tasks (survey, refine, map and sample) are achievable. Phase-2 challenge environments provided by Dstl will start with simple indoor (single-room, minimal internal structures, still air), simple outdoor (the approach to the building entrance), and simple transitions between these environments (entry assured to building); more complex indoor (2 storey house, complex internal structures, including manufacturing areas) and much more challenging outdoor vignettes will then follow, involving a wider scale of contamination(for example, in a complex urban site with multiple surfaces). Phase 2 may also include land, air and maritime platform environments.

7. Important information

This 2-phase competition will be supported by presentations given at the CDE Innovation Network event on 22 September 2016 and at a webinar on 3 October 2016.)

Your phase-1 proposal must be received by CDE by 5pm on Thursday 10 November 2016. Proposals must be submitted to CDE online.

We won’t accept phase-1 proposals over £100,000 and it’s more likely at this stage that a larger number of lower-value proposals (eg £40,000 to £80,000) will be funded than a small number of higher-value proposals. Total funding available for phase 1 of this competition is £1.5 million.

Read important information about how to submit a proposal to CDE.

Read important information about CDE contracts here. For this competition you will also need to take into account that DEFCON 705 (Edn 11/02) shall apply with the following clarification:

The “Authority” as referenced within this Condition at clauses 12, 13, 30, 31 shall be deemed to be any United Kingdom Government Department.

Proposals will be assessed by subject matter experts from MOD, Dstl, Home Office and wider government. Read about how your proposal is assessed.

If you’re successful and awarded a contract, Dstl will appoint a technical partner to provide you with advice and/or guidance throughout the project and act as the interface with Dstl, MOD and wider government stakeholder community.

Deliverables from contracts will be subject to review by MOD and the Home Office.

This competition is made up of 2 phases.

Phase-1 proposals should focus on an initial short, sharp, proof-of-concept phase up to 6 months in duration, with proof-of-concept research completed by 30 June 2017. In Phase 1 we’re exploring applicable technologies and systems to address the defined tasks. We expect to fund a number of projects to run for up to 6 months to explore a range of possible system elements and solutions, from component level through to integrated systems. Successful bidders from phase 1 will develop and demonstrate their technologies in a well-controlled laboratory or workshop environment. Your proposal should clearly state the metrics you’re proposing to use for this demonstration.

Your phase-1 proposal should include a descriptive scoping of the next phase of the programme (phase-2), but the proposal should be clearly partitioned with a fully costed proof-of-concept for phase 1, which is the focus of this CDE themed competition.

A deliverable of your phase-1 proposal will be a fully costed proposal for phase 2 work. Phase-2 of the competition is only open to projects successfully funded at phase 1 and proposals will only be considered after the phase-1 proof-of-concept has delivered. The understanding gained in phase 1 will allow Dstl to make an informed decision about funding for future work. Total funding available for phase 2 of the competition is £1 million and proposals will be submitted via the CDE online submission service.

8. Materials of concern and use of simulants

Although we’re interested in assessing scenes contaminated with very dangerous materials, you won’t need to worry about using them in this competition. However, we expect to see evidence that the final systems delivered at phase 2 have well-defined, direct growth paths to address highly toxic chemicals and/or dangerous pathogens, and this should be described in your final submission at phase 2.

But for the purposes of this competition you must use simulants. These are materials that mimic highly-hazardous materials sufficiently to demonstrate your discoveries and developments, but that are far safer to use. For instance, simulants materials will be required to trigger responses in sensor payloads that will form part of the systems at phase 2, and where applicable in your work at phase 1; similarly, they will also be required to demonstrate that sampling technologies are working as intended.

In your proposal you must clearly define the simulants against which you’ll demonstrate your technologies’ benefits wherever they’re required, at both phases of the competition. We don’t want to constrain your technologies to specific simulants, as a trigger for one technology may not be appropriate for another. You can decide which simulant materials you’ll use but they should have the following properties:

• some materials of interest are liquids that have relatively low vapour pressure under standard conditions, which results in a significant amount of residual liquid post-release. Others have a much higher vapour pressure, which leaves less residual liquid but generates a larger volume of vapour. The former may have boiling points in excess of 250 degrees Centigrade, whilst the later may boil well below 50 degrees Centigrade
• other materials of interest generate solid or liquid aerosol particles. Some have a significant mass fraction in the 1-10 micron mass median diameter region, and are predominantly air-buoyant but may also settle onto surfaces. Others have significant mass fraction in the 100 micron-1mm region, leading predominantly to surface deposition, but with potential of re-aerosolisation disturbed. And finally, others fall into the intermediate case (significant mass fraction in range 10-100 micron).
• airborne mass contamination concentrations might range from µg/m^3 to g/m^3
• distribution of bulk material might vary dramatically - from random droplet patterns to thin continuous films and even puddles.
• surface mass contamination densities might range from mg/m^2 to kg/m^2

You should also appreciate that Phase 2 demonstrations will be conducted in a variety of environments – some will be indoors with still air environments; some will be outdoors, and experience ambient conditions on the day of demonstration. Depending on the time of year that these demonstrations take place, temperature could vary from 5-35 degrees Centigrade, across a range of relative humidity.

9. Ethical considerations and The Regulation of Investigatory Power Act (2000)

9.1 MOD Research Ethics Committees

All research involving human participation conducted or sponsored by MOD is subject to ethical review under MOD procedures as outlined in Joint Service Publication 536 ‘Ministry of Defence Policy for Research Involving Human Participants’, irrespective of any separate ethical procedures (eg from universities or other organisations). This ensures that acceptable ethical standards are met, upheld and recorded, adhering to nationally and internationally accepted principles and guidance.

The following definitions explain the areas of research that require approval:

• clinical: conducting research on a human participant, including (but not limited to) administering substances, taking blood or urine samples, removing biological tissue, radiological investigations, or obtaining responses to an imposed stress or experimental situation

• non-clinical: conducting research to collect data on an identifiable individual’s behaviour, either directly or indirectly (such as by questionnaire or observation)

All proposals should declare if there are potential ethical issues.

Securing ethical approval through the MOD process can take up to 3 months. In this CDE themed competition, projects must be completed by 30 June 2017 and obtaining ethical approval could take your proposal beyond the timeline for completion of phase 1. We, therefore, recommend that you only include research in phase 1 that doesn’t require ethical approval. Work that might require ethical approval should be planned for future phases of work which are likely to have longer and more flexible timescales.

However, if you think that your phase-1 proposal may require ethical approval, please ensure that you take an approach in your submission as follows (noting that projects must still complete by 30 June 2017):

• milestone 1: gaining ethics approval for the project, including delivery of the research protocols (the protocol will need to be detailed by completing the ethics application form)

A contractual break point should be included after milestone 1.

Read more on the MOD Research Ethics Committees.

The requirement for ethical approval isn’t a barrier to funding; proposals are assessed on technical merit and potential for exploitation. Successful proposals will be supported through the ethical review process; however, an outline of your research methods must be included in your proposal to help this process.

9.2 The Regulation of Investigatory Power Act (2000) considerations

The Regulation of Investigatory Power Act (RIPA) is an act of the Parliament of the United Kingdom, regulating the powers of public bodies to carry out surveillance and investigation, and covering the interception of communications. You should ensure that your proposal or the work you’re proposing to undertake doesn’t cause a breach of these regulations.

10. Dates

22 September 2016	Competition briefing at Innovation Network event
3 October 2016	Webinar
10 November 2016	Competition closes at 5pm
January 2017	Contract placement initiated and feedback provided
February/March 2017 (date to be confirmed)	Collaboration networking event
May/June 2017 (date to be confirmed)	Collaboration networking event
30 June 2017	Latest date for the delivery of phase-1 proof-of-concept research
July 2017 (date to be confirmed)	Phase-1 stakeholder event
8 August 2017	Phase 2 competition closes at 5pm for submission of phase-2 bids
Mid-September 2017	Phase-2 funding decisions made
October 2017	Phase-2 contracts placed
October 2018	Latest date for delivery of phase 2 projects

11. Queries and help

While you’re preparing your proposals, you can contact us if you have any queries:

Technical queries about this competition should be sent to: dstlautonomyinnovationchallenge@dstl.gov.uk

Capacity to answer these queries is limited in terms of volume and scope. Queries should be limited to a few simple questions or if provided with a short (few paragraphs) description of your proposal, the technical team will provide, without commitment or prejudice, broad yes/no answers. This query facility is not to be used for extensive technical discussions, detailed review of proposals or supporting the iterative development of ideas. While all reasonable efforts will be made to answer queries, CDE and Dstl reserves the right to impose management controls when higher than average volumes of queries or resource demands restrict fair access to all potential proposal submitters.

General queries should be sent directly to CDE at: cde@dstl.gov.uk