a whole exoskeleton is hard due to the portable energy density limits ... I've seen some preliminary work from liquid-prosthetics coming out of academia but not to the point of a spinout. I attended a funding event where a US company was pitching to expand some of their manufacturing operations to china but that wasn't a whole solution. The market may just be a little immature which is good for hackers and toy-tinkering (akin to Apple 1 days) but harsh for serious startups. Defence and medical prosthetics as always been a specialty market, there's the whole prospect of retiring baby boomers for assistive technology but for mobility purposes, scooters/segways give better power/distance to utility ratios. I know the Japanese are working on domestic robots but it's a different order of engineering for exoskeletons.
Let's consider the fundamental issues - strength to weight, power density, cybernetic feedback times. For anything to be worn by disabled, it needs to be closer to Ironman than MechWarriors, which implies low weight/bulk but decent structural integrity which implies advanced materials which is not cheap. Suddenly you've cut your prospective market to top 10% or 1%ers who can probably afford personal aides anyway. For something useful it needs to work over an interval to justify getting in/out which means external power ... if a phone just after a few hours just imagine the energy storage to run a skeleton for a large chunk of the day. There's some work on printed lithium batteries which can be form-fitted but the weight becomes more of an issue which is why armies don't carry around armored suits anymore. Then look at the challenges of control, interfacing with nervous system and balance. Our brains have been trained from childhood to intuitively map spatial cognition with muscle reactions and sensory feedback. Thus becomes an individual tuning process and even with AI techniques is non-trivial.
So from Australia's PoV, materials science is hardcore but has spinoffs in autonomous vehicles as well as personal mobility, power technology is engineering, but cybernetics is probably doable. Whether a startup can be sustainable depends on what contracts they get with the hardware side. I believe there's some people in Perth working on underwater autonomous vehicles and probably some industrial groups on automating the mining and shuttle tankers long-distance carriage problems. However, without access to a decent high density power sources, I conjecture personal exo-skeletons are going to be hard to get off ground in Australia.