The refinement and application of autonomous systems is accelerated at an exponential rate that changes the way we move, work and interact with the world and those around us.
Even this article was brought with the help of transcription services from speeches that use machine learning technology to transfer spoken words into text in real time. And while it's not perfect for accurately interpreting language, it does more than a stunning job of documenting arguments and summarizing key points.
As humans become more dependent on autonomous systems and artificial intelligence, how can we make machines, conscious or not, more reliable and make their interactions more human-centered?
How can we ensure that self-driving cars and planes respond to instructions and take us safely to our destination?
Or does it mean that support robots for the elderly and disabled can communicate with clients in a more human-like way?
Human Machine Collaboration
Enhancing this human-machine collaboration is the focus of lot 14 research between the US non-profit research and development organization MITER and the University of Adelaide's Machine Learning Institute (AIML). AIML operates on the cutting edge of AI Research and has developed several pathways to promote the transfer of that knowledge between Australia and the US.
The study is based on a partnership signed by Miter in 2022, and the University of Adelaide focuses on building Australia's AI capabilities for future security and prosperity.
“By developing and enhancing the capabilities of autonomous systems such as robots, drones and self-driving cars, we can help us interact with the world in a more human-like way to interpret the environment using sensory data.”
“This will allow the system to be more intuitive, safe and able to work with humans, helping to improve both operational efficiency and quality of interaction with people.
“We're not there in any way. I hope this research will help solve that challenge.”
The 18-month research project will be carried out at AIML in Adelaide.
Embedded AI and Robotics
The research is led by Dr. Feras Dayub, an expert in AI and robotics. Dr. Dayub's work is dedicated to promoting the reliable deployment of computer vision and machine learning on mobile robots in real-world environments, ensuring that they understand the situation and communicate appropriately.
“The findings from our research using Miter are to explain what it is doing, what it did, what it didn't do, what it was trying to do, and provide that information in a more natural way that users understand,” says Dr. Dayub, a senior lecturer at AIML.
“When a robot is deployed in real terms, sometimes things don't go as planned. There are many variations. [and] The method of sensors can cause obstacles and environments to change, which can lead to tasks failing.
AIML performs experiments in the laboratory with a robot, DART, or dual-arm robot (pictured). The AIML and MITRE partnership includes knowledge and data transfer, allowing AIML research to be tried out in sophisticated autonomous systems in MITRE's extensive laboratory and research facilities in the US.
State-of-the-art testing facilities
The MITRE MASE (Interdisciplinary Autonomous System Assessment) Lab in Virginia, USA has Jeep equipped with sensors, analytical and data recorders, and powerful computer processors to test and document new autonomous technologies. The facility is designed to test real challenges such as sensor malfunctions and unexpected failures.
Miter also has a state-of-the-art indoor maritime testing facility called Miter BlueTech Lab in Massachusetts, and Miter National Range in Virginia is dedicated to testing private aircraft systems.
“It's worth taking the software, putting it in the hardware and doing something really cool,” says Wotton.
By testing AIML frameworks in these labs, MITRE can identify vulnerabilities before being deployed into the field, fine-tune system performance, and ensure that autonomous systems can perform performance safely and predictably under a variety of conditions.
“The findings of the research will enhance the AI capabilities of both the organization and its government partners,” says Wotton.
“Our research is driving innovative and cost-effective solutions to government challenges, from national security and transport safety to cyber defense,” he says. “We share what we discover. It's part of our public interest mission.”
The company's technology transfer program places innovative solutions in the hands of governments and industry, as well as the global security community and academia. This knowledge sharing commitment encourages public and private sector impact.
“The AIML-MITRE partnership will allow us to jointly migrate technology solutions and prototypes that could lead to improved safety, and ultimately save lives,” adds Wotton.
“If a research announces a potential vulnerability, it can be incorporated into various published open source frameworks from Miter, such as Miter Att & CK® and Miter Atlas.
Future research applications
- 1. Advanced Aviation and Transportation: Fully autonomous vehicle, intelligent AI co-pilot, route planning and optimization.
- 2. space: Autonomous navigation and exploration, robot mining.
- 3. health: Surgical robotics, assisted robots for the elderly and disabled, medical triage, logistics and delivery, precision medicine.
- 4. Advanced Manufacturing: Workforce empowerment, process optimization, high-precision automation, smart warehousing.
“By partnering with an organization like AIML, we can move some of our key research to Australia and develop a kind of sovereignty capacity while MITRE has been developing for over 65 years,” says Wotton.
“Our goal is for AI-enabled systems to be safe, secure, reliable and benefit end users around the world. This collaboration is key to moving forward and sharing AI promises.”
