Autonomous cars & speed limits

The progression of electric vehicles is being mirrored by the progression of autonomous vehicles, with most mainstream manufacturers now offering varying levels of driver assistance and autonomy. SAE Level 2 autonomy enables "hands off" driving, where the vehicle maintains speed, braking and steering, whilst the driver remains "in control" and ready to take over should the vehicle not choose the correct course of action. Essentially any vehicle with adaptive cruise control and lane departure warning technology is knocking on the door of Level 2 autonomy.

This year the UK government gave the go-ahead for the first trials of platooning vehicles, which is expected to happen later in 2018. The significance of this trial is not in the technology; this has been proven in the lab and on closed tracks and there is no way that the public highway would or should be used to find out if technology works. The significance is to characterise the known unknowns, especially how other drivers and vehicles on the road interact with the platooning vehicles, and the emergent behaviours this will create.

There appear to be two main objections to platooning vehicles on UK motorways; firstly that they'll obstruct slip roads, and secondly that they'll obscure signage. On a personal level I disagree with both of these points. Slips roads are signposted a mile in advance, giving vehicles plenty of time to filter into the left hand lane, and bigger vehicles can already obscure signage from smaller vehicles, so this is not an issue caused by platooning vehicles.

However the huge significance of this trial for me is that permission has been given for platooning vehicles to break rule 260 of the Highway Code, i.e. "...keep a safe distance from the vehicle in front". In this trial the platooning vehicles will not be a safe distance from each other, if they were being driven manually. And this is an important point, as platooning vehicles are operating at Level 3 autonomy, where the driver is not able to take control should the vehicle not choose the correct course of action. At Level 2 autonomy the vehicle is always being operated safely, as the driver is ready to take over if the autonomous systems are not operating the vehicle safely.

So this trial is setting a precedent whereby a Level 3 autonomous vehicle need not meet all the rules of the Highway Code. Rules 124 and 125 immediately spring to mind, as they deal with speed limits. The key to this debate is the unbalanced nature of these rules, especially as captured in rule 125: "The speed limit is the absolute maximum and does not mean it is safe to drive at that speed irrespective of conditions." I don't think many people would argue that in the depths of a snowy winter driving exactly at the 30mph limit down a street crowded with festive shoppers may not be not safe. But I also don't think many people would argue that exceeding the speed limit by a few mph on a dry, deserted motorway is any less safe than at the speed limit.

So the debate that is looming is this: if a Level 3 autonomous vehicle can determine when it is safe to platoon, surely a Level 3 autonomous vehicle can also determine when it is safe to exceed the current speed limits, and by how much. And this will be a debate based on opinions and feelings rather than cold hard facts. There is hard data to prove that autonomous systems can safely drive a car well in excess of the speed limit, but many drivers are still wary of much simpler driver aids. For example I know many experienced drivers who shun manual cruise control. These are the people that need to be won over.

Hopefully they will be won over, because higher levels of vehicle autonomy are key to free-ing up the UK's motorway network. Instead of smart motorways with variable speed limits and roadworks to add lanes to already congested roads, we could have more cars, travelling faster, closer together, and all safer than we are at the moment. What's not to like about that?

WiFi Coverage - Stage 2

Today I've been playing with setting up WDS (WiFi Distribution System) on the two DrayTek VigorAP 910C access points that beam the internet to the workshop. In principle this should enable them to function both as a point-to-point link between house and workshop, and function as access points for wireless devices nearby. For wireless devices in the workshop, this would mean that they could connect to the 910C in the workshop, and have that wireless traffic relayed to 910C in the house, and then onto the router and the wider internet.

Previously I'd used two different SSIDs for the networks on the 2.4GHz and 5GHz bands, as I found it useful to know which band a device was connecting on. But the DrayTek website makes such a big thing about keeping the configuration of the 910Cs exactly the same, even down to the wireless channel they're using, that I decided to set up all of the SSIDs and pre-shared keys on each 910C the same. This would also enable them and the Vigor 2860ac router to use band steering, and bump any capable wireless devices onto the 5GHz band automatically. This in turn would keep the 2.4GHz band as clear as possible, which should all help with the bandwidth of the point-to-point link with the workshop, which is only in the 2.4GHz band.

Setting up the 910Cs for WDS was fairly straight forward in the end. After setting the Operational Mode to AP Bridge - WDS, the settings and rules used for Point-to-Point mode transferred over. But whereas previously wireless devices at the workshop couldn't see an access point, they can now. Unfortunately it doesn't appear that in WDS mode the 910C can internally route traffic from wireless devices connected on the 5GHz band to the 2.4GHz WDS link, just like the point-to-point mode. So there's still a hole in my WiFi coverage at the workshop, in the 5GHz band. Looks like that trench is inevitable.