Unmanned Systems Technology 028 | ecoSUB Robotics AUVs I ECUs focus I Space vehicles insight I AMZ Driverless gotthard I InterDrone 2019 report I ATI WAM 167-BB I Video systems focus I Aerdron HL4 Herculift

37 for electrical generation that is directly proportional to propulsion power. It is also likely in the years ahead that ECUs will increasingly be expected to incorporate greater autonomy (via longer and more complex decision- making algorithms). For example, by intelligently understanding the nature of its mission, a UAV could respond to a minor engine fault by determining whether it can continue ascending or loitering, how it might improvise to achieve its objective without exacerbating the problem, or how long it can postpone its return flight before the fault becomes critical. Improvements in engine health monitoring and data logging are likely to continue, towards the widely touted goal of AI-based predictive maintenance based on analytics software, which is designed to anticipate engine failures tens or hundreds of hours in advance. Engine and ECU developers will also continue to see higher demand for a heavy-fuel capability, which requires more sophisticated engine control strategies than ECUs for typical gasoline engines. The ability of ECU programmers to quickly access and modify their software stacks to implement complex new requirements will be key in this regard. Additionally, there will almost certainly be future step increases in demands for electromagnetic insulation in ECU enclosures. This is partially due to the need to reduce EMI between the ECU and other RF-emitting onboard systems, such as electric motors, navigation sensors, data links and so on. However, it will also improve immunity to electronic countermeasure systems, jamming and interference from high- power ground-based transmission infrastructures communicating with satellites and other off-world structures. It was still common to use automotive ECUs as the basis for new UAV units as recently as three to five years ago. However, automotive designs fall short of the latter in ways that the market may not now tolerate – not least because they are not weight-optimised in the ways UAVs need. For example, when it comes to pressure readings, an automotive ECU will read 50 kPa of pressure as an error, because no road in the world goes above 4.25 or 4.5 km – but UAVs do. Furthermore, UAV engines are used to drive fixed- or variable-pitch propellers, with remarkably different thermal management and gearbox systems than a car. Similarly, ECUs must have built- in fault tolerances, and these also differ wildly between cars and UAVs. Temperatures and cooling For example, a cylinder head temperature sensor might need its ECU to be programmed with over-range and under-range temperatures for identifying anomalous levels of heat, as well as for corresponding cooling output commands. Since cars and UAVs operate in different environments with different engine architectures, however, the operating temperatures and cooling requirements will be completely different. Additionally, many small UAV engines (particularly two-strokes) do not have real-time exhaust gas feedback, and the absence of data feeds such as this could be read as critical errors by an automotive ECU. Fundamental programming and architectural differences such as these make it more sensible to start from a blank sheet than repurpose an automotive ECU for a UAS. To custom- design an ECU from scratch, or to adapt an existing UAV ECU architecture to a new engine or vehicle, a range of application requirements must be included from the outset. For one, the environment in which the ECU will operate must be fully understood. This includes accurate estimates of altitude, shock, vibration, humidity and so on. That will establish the requirements of the enclosure (almost always the heaviest part of the ECU). Engine control units | Focus Unmanned Systems Technology | October/November 2019 ECUs must handle a broad range of inputs and outputs, which is certain to widen as hybridisation and increasingly complex payload systems draw on engine power and torque generation (Courtesy of Moscat Ingenieria)