Still on the motor theme, one big thing on the horizon for us is effective sinusoidal filtering of variable speed drives. As Skogsgurra says, PWM stray currents can be destructive and filtering is one solution - the current batch of filtering technology on the market have various trade-offs that depending on the application can cause more dramas than they solve. Newer designs - NFO Sinus, current-source topologies, etc., could have a huge impact if they scale well.
On the electronics design front, Functional Safety design is a lumbering elephant on the horizon. Perhaps already pervasive in aviation and automotive, "Proven" operating systems, redundant architectures, voting systems, quantitative FMEA, formal specification languages, all have the potential of fundamentally changing the way we operate.
And on the silicon level, massively parallel architectures will be one of the next big things - chips are already being manufactured with hundreds of parallel processing cores, but the software and algorithms to get the most out of them is still playing catch-up. As we approach physical limits in clock speeds, parallelism will be the next way forward.
