The Case for Response?Robotics

For many years I have expressed my disappointment with the lack of technology diffusion in Australia. In particular from the mining industry - where we have led the world in field robotics and remote operations for a number of decades, to other sectors such as Agriculture, Construction and Defence. I have attributed this to culture of commodities and inbound innovation .

This is the topic of my article on "Innovation in Australia "

At the same time, I have been very interested in working with Australia's Defence industry - both with the Australian Defence Force (ADF) and the Defence Science Technology Group (DSTG) - but the demarcation between civilian and military research can be fuzzy and difficult to navigate.

But in 2019/2020 everything changed.

• We were hit with the triple whammy - fire, flood and pandemic.
• The world entering into an era of continuous disruption.
• And we are going to need a lot more technology - more D-Tech

I was called to participate in CSIRO's Bushfire Working Group. From these meetings it became apparent that no one was thinking of robotics as a potential response to bushfires. Yes drones could be used for surveillance, but there was no discussion on how robotics could physically respond to the fire - to intervene - to actually fight the fire.

To raise awareness of the capability of robots to act, I hosted the Response Workshop of the 2nd Robotics Roadmap for Australia (see slides on Bushfire Use Case ). From this workshops it soon became apparent that there was a great deal of interest, and some interesting dynamics and synergies between the mining and defence sectors, including the possibility that disaster tech could act as a stepping stone between the two - catalysing the diffusion and translation of innovation across Australia.

But we have a number of problems in Australia

1. There is very little money in disaster technology. Whilst, there may be plenty on offer immediately after a disaster - this can dry up (pun intended) very quickly. There is no sustainable longterm R&D funding.
2. Procurement in defence is primarily interested in proven technology. This can be very difficult and expensive to do in Australia, so we source most of our technology from OS.
3. The mining industry has been very wary of being directly involved in defence - it can compromise the ability to export technology and their social licence.

But we have a number of advantages:

1. Austrade have identified D-Tech as a significant export opportunity [1]
2. The Mining industry has shown an interest in supporting D-Tech [2]
3. The Defence industry is now very interested is supporting D-Tech [3]
4. Both industries are interested in developing a Sovereign Capability [4]
5. Two new technology networks were formed - AROSE and RAN

So by placing disaster between mining and defence we are able to develop technology that has both civilian and military applications - called Dual Use Technology - and when required, building a Chinese wall between the two sectors.

So what is Disaster Tech?

Disaster tech is any technology that can reduce the impact of a disaster - save life or property. The technology can be used in all three phases of a disaster: Before - Resilience; During - Rescue and Relief and After - Recovery. The disaster can be single event - such as a flood or fire, or it could be extended over a period of time - pandemic or coral bleaching .

Disaster tech can range from data management, simulation, communications, displays, wearables, devices and machines - including response robots. An example is SPARK , a fire prediction platform developed by my colleagues in CSIRO for bushfire modelling that integrates fire weather data with geographic information and fire spread models.

So what are Response Robots?

Response robots are the set of robots that span the three phases of a disaster: from resilience, rescue and relief, to recovery. They can also include robots that deal with events at a different scales: from a small incident, to an emergency, to a full scale disaster. And the roles can range from surveillance to intervention. This is mapped out in the figure below for bushfires: starting with monitoring and management of ecosystem (fuel load); to the actual event, which could be a spot fire, a localised bush fire, or a statewide bushfire; back to monitoring and management of the ecosystem for rehabilitation and recovery.

Within the robotics community all of these robots lie within the classification of Professional Service Robots - which is the subject of my article "The Next Robotics Race is On ". In the figure below are some response robots developed here in Australia

What makes a response robot?

When people talk about robotics they usually mention the traditional 3D's - Dull, Dirty and Dangerous. But there are 3 extra D’s for Response Robotics (UKRAS) which includes:

• Demanding (i.e. beyond the skill of direct human control);?
• Distant (i.e. avoiding the logistics problem of having to transport human workers);?
• Distributed (i.e. allowing agile and scalable deployment).?

This last point is one of the most significant and overlooked feature of response robotics. What robotics truly offers is the ability to be deployed at a moments notice (agility) and at scale - to act as a force magnifier. If you build and deploy one robot, you can technically deploy a 100, or a thousand - swarm robotics [5].

Response robots operate in:

• Unstructured dynamic and hostile environments
• Without supporting infrastructure
• Reconfigurable, Redeployable and Resilient or Redundant

Whilst there are many challenges - the most significant is the business case.

How can we pay for all these robots?

Conventionally, we think of the three phases of a disaster as three different problems, requiring three different solutions, and hence three different robots, but let us consider a situation where the robots are used in all three phases, and the mission of the robot is simply reconfigure for each phase.

During Covid-19 we have learnt a number of lessons [6]

1. Robots should help workers, not replace them
2. Deploy off-the-shelf robots not prototypes
3. Don’t stockpile robots - they rot.
4. We can’t build a robot for every event!
5. We need to create community of capability (skills)

Response Robotics requires a Change in Thinking

One of the challenges with robotics is the way they are portrayed in the media.

When we hear reports of robots/drones being used in emergency we see an image of a first responder with a remote control in their hands. This is not how we should see response robots. The robot should be seen as relatively autonomous - being monitored from a mission control. It requires new thinking

Response Robots of the Future

So with this new thinking, imagine a scenario where we have 1000's of drones (airborne, surface and submarine) patrolling the coast of Australia. Theres drones could be repairing the reef, monitoring fish stock, conducting geological mapping, removing plastic etc etc. And then, with one command, the drones could be activated into defence mode.

Sort of like the red light in "I robot".

This is the reverse of the recommendations of the Royal Commission into bush fires - where Defence would be called upon assist with disasters. What we have here is the ability for Defence to conscript civilian robots, into defending our country from external aggressors or natural disaster. This can be extended even further. Imagine when all of roads, farms and mines are populated with robots. Could all of these robots be called upon in an emergency.

Of course, cybersecurity would then be of paramount importance.

References

[0] 2020 CSIRO Report on Climate and Disaster Resilience

[1] 2020 Austrade Global Trade and Investment Megatrends report -

Action 8 - Developing an export-earning disaster-resilience technology industry. Global demand for trusted technologies for the management of wildfires, droughts, heatwaves, floods, pandemics, and cybercrime is set to escalate. As we develop solutions to these challenges in Australia there’s an opportunity to identify, and supply into, global export markets. In summary, this strategic action involves building Australia’s brand, reputation, and capability for the supply of trusted solutions to disasters worldwide. It also involves developing market intelligence to identify buyers and their specific requirements.?Key opportunities for Australia is to Develop an export-earning disaster-resilience technology industry

[2] 2020 Minderoo Foundation -?Fire and flood resilience

[3] 2020 Royal Commission - National Natural Disaster Arrangements

Acknowledgement that the ADF play a more significant role in Disaster Management - Defence and Disaster coming together ?

[4] Private conversations with Joint Capabilities Group - about the development of Dual Use Technology by Australian Manufacturers to support sovereign capability, Conversations with RAN (Autonomous Warrior), Army (HumT) and RAAF (Jericho) on the need for domestic proving grounds for response robotics. Conversations with DMTC, AFP, AGO, DST on technology for first responders.

[5] 2021 ARMY - Report on Applied Research Directions and Future Opportunities for Swarm Systems in Defence - Adam Hepworth

[6] 2020 The Conversation Robots-are-playing-many-roles-in-the-coronavirus-crisis-and-offering-lessons-for-future-disasters

The broad use of robots for COVID-19 is a strong indication that the health care system needed more robots, just like it needed more of everyday items such as personal protective equipment and ventilators. But while storing caches of hospital supplies makes sense, storing a cache of specialized robots for use in a future emergency does not.\

International References

• NIST (US National Institute for Standards and Technology) - Response robots
• UKRAS (UK Robotics and Autonomous Systems) Network - Extreme Environments Robotics: Robotics for Emergency Response, Disaster Relief and Resilience
• Japan DDT -?Robots and Systems for Urban Search and Rescue

Domestic References

2020 Chief Scientist Bushfire Research and Technology Capability Map

2021 InnovationAus Space, robotics among new Defence priorities

2021 Defence Sovereign industrial capability priorities

2019 NERA and Metsignited - Staying ahead of the Game

2021 Department of Industry Australian Artificial Intelligence Action Plan 2021 - Australia is already a leader in strategic fields of AI like computer vision, deep learning, field robotics, neural networks and machine learning.?

2018 Department of Home Affairs Disaster has Emergency Management Australia and National Disaster Risk Reduction Framework

2010 Rescue Robot Breaks down - Australian Mining

2010 Rescue Robot Breaksdown - Efforts to rescue 29 men trapped in a NZ coal mine suffered another setback when a robot sent into the main shaft broke down

2020 Robotic camera and drone fail to find missing Tasmanian mine worker - Police, fire and mine rescue crews have not located a man missing inside a gold mine on Tasmania's west coast after an earth collapse in the early hours of Thursday morning

2022 Army Research Centre - Robotic & Autonomous Systems Strategy v2.0