Over the last week or two there has been lots of debate about unnamed 'technological solutions' for avoiding a hard border in Ireland. So what are these mooted solutions? 

Firstly, we need to establish what this type of border is usually used for, taking into account pedestrians, cars, freight and trains. Air travel is not as pertinent as there are already measures in place. Taking these in turn:

1) Pedestrians. The increasing sophistication of facial recognition technology, and its interface with AI to quickly analyse multiple databases, provides an initial possibility.  Live examples can already be found at various major airports.

2) Goods transport. A lot of this could in theory be dealt with before reaching the border - currently, much freight documentation is dealt with digitally. By entering the roster details onto supply-chain software (that could leverage blockchain technology) and potentially even by using gamma/x-ray scanners to monitor cargo as it passes through, we could avoid the need for manual checks. 

3) Motor transport. This gets slightly more tricky because the facial recognition camera may not be able to accurately see all the faces in the car. Perhaps if there were enough cameras in different positions, coupled with Automatic Number Plate Recognition (ANPR), GPS phone tracking and preregistering your passport online (or a block-booking for regular commuters), it might work. 

4) Trains. We could again look at facial recognition onboard trains and preregistering passports, coupled with biometric/fingerprint scanning at stations. 

I think the overriding theme here is that we are not actually avoiding a border at all - it will only be a case of making the border as invisible as possible, which is known as 'policing at a distance'. As referenced in the quoted article, some countries are already implementing borders like this (e.g. Norway), but the chances of this being implemented by the transition period deadline in December 2020 appear slim. 

In conclusion, we can see that there is some technology available that could form a smart border toolkit (even with a fair bit of speculation), but whether it works in practice certainly remains to be seen - and that is not considering the substantial costs involved.