UST030

98 PS | Lifespan of unmanned vehicles H ow long should unmanned vehicles last? Some are inevitably disposable to some extent, particularly military ones developed for the dull, dirty and dangerous task set, but our culture is becoming increasingly uncomfortable with disposability in a broadly sense – particularly when expensive products smell of planned obsolescence (writes Peter Donaldson). There’s more than a suspicion that many high-end cars for example are designed to be reliable for the first owner, who will most likely keep them for three years before moving on to a new one. After that the cars start to be very expensive to maintain, and depreciate heavily. This apparent disposability even seems to affect agricultural machinery, perhaps significantly, with cab parts handled by operators tending to fail first. An extreme counter-example comes – ironically – from the military. The US Air Force is still operating more than 70 B-52 bombers that are older than most of the pilots flying them. The last one was built in 1962, but diligent maintenance, repair and overhauls have kept them flying, while numerous updates and fitting new technologies have kept them operationally relevant. For many engineers, working on something potentially as long-lived as that is very satisfying, and indications are that many in the industrial and professional unmanned systems sector are thinking in this way. At least two of the vehicles discussed in this issue emphasise the use of highly durable materials and easily repairable subsystems for reliability and longevity. While agricultural vehicles have to be agricultural, that adjective does not have to be disparaging. GUSS Automation for example (see page 92) was determined to build the best orchard sprayer it could, one that could last for decades. Similarly, Griff Aviation (see page 52) emphasises its use of CNC-machined aluminium alloy for the airframes of its heavy-lift multi-copters, with aesthetics and perceived quality high on the list of priorities along with durability and longevity. A logical extension of this is building machines that can maintain and repair themselves. Researchers at the George Washington University (GWU) for example have applied real-time thin-film sensors and nanotechnology to a system that detects surface damage in moving parts such as gears, and repairs them automatically. As GWU researchers under Professor Stephen Hsu are also working on micro- encapsulation and controlled release of complex chemicals in lubricants, the repair process probably involves the release of such chemicals at the damage site. Meanwhile, the Self-Healing Soft Robot project led by the University of Brussels is applying self-healing elastomers to components such as grippers and pneumatic actuators that are subject to cuts and perforations, using machine learning to integrate them into controlled self-healing processes. Also, engineers at New York’s Columbia University, led by Professor Hod Lipson, have demonstrated an arm that, using random movements and deep learning, created an accurate model of itself and learned to do pick-and-place tasks. When fitted with a deformed component, it proved able to sense the change and compensate. Nothing lasts forever, but by valuing quality and longevity, and developing self-repair, engineers can show entropy some defiance. Now, here’s a thing “ ” Many engineers of unmanned systems are looking at ways to make them as long-lived as the US Air Force’s B-52 bombers February/March 2020 | Unmanned Systems Technology