It really is astounding at how we can so easily erase the need for basic pedestrian amenity when new technology rolls around. Even the people at Amazon did not see the obvious usurping of the sidewalk by their six wheeled  sidewalk delivery robot as being a problem. You can see in the video below as “Scout” (yes they named him) dutifully takes up most of the sidewalk as he rolls on his route. There is no space left over for a pedestrian, a baby buggy or a user of any mobility device.

But now Amazon as reported in the Business Insider wants pedestrians to know that not only do they need to give way for these robots on narrow sidewalks, but “that the public should treat these robots in the same way that they would pedestrians.”. 

Yes you heard that right. The sidewalk delivery robots want to have the same rights and the same rules of the road as do pedestrians, and also will use the road only if a pedestrian would do the same with a level of comfort.

Sean Scott of Amazon states “If you feel safe walking on that road, that’s where we want to be. We want to be viewed as a pedestrian and treated as a pedestrian.”

So here is one more device that wants to take sidewalk space without paying any charge or user fee for taking that away from mobility aid users and pedestrians. And remember autonomous vehicles? Their big hiccup has been the inability to distinguish objects on the road,  and the unpredictability of pedestrians. You can now add sidewalk delivery robots to that mix.

You can take a look at Amazon’s write up of how Scout will work and marvel once more at the ability of technology to see an end use to their innovation, with no concept that sharing the road or sidewalk with these devices may come with a cost for the upgrade and maintenance of these facilities. It should not be a  free ride for Amazon.


  1. lol. In the sometimes rough-and-tumble neighborhood I grew up in many years ago, that buggy wouldn’t get 10 feet. I wonder if it would be different now? I know my dogs wouldn’t like it.

  2. Unlike self-driving cars, I this could really work in urban environments. I think that’s bad. We badly need more sociable, human cities. We don’t need more convenient consumption, and we certainly don’t need it taking over our sidewalks. I hope our civic leaders get ahead of this before it becomes entrenched.

    Speaking of which, e-scooter rentals have arrived in Burnaby. I think they could be wonderful if done well – i.e., if we pay attention to problems elsewhere. My relatively uninformed suggestion is that they should be designed with a top speed of, say, 12-15 km/h, and that drop-off should only be permitted in specific areas (e.g. the walkways between sidewalks and bike paths along town centre roads, but not on the sidewalk or bike path itself). If they work out, we should probably set up drop-off areas at Skytrain stations.

    1. Not quite sure why an autonomous buggy would work but not an autonomous car. Ultimately, what’s the difference? It’s completely unacceptable for these things to go around hitting people even if it’s unlikely to kill them.

      1. In either case, the vehicles would hit people from time to time. I think it’s completely unacceptable that ordinary cars kill so many people – but they do. What’s unacceptable is ultimately a political question of what people will actually rise up and reject.

        The issue, it seems to me, is the rate of errors (leading to collisions) and the consequences of those errors. A sidewalk vehicle travels at low speeds. People can avoid slow-moving objects (like zombies on phones) almost without thinking. It has a shorter stopping distance (effectively zero I expect); is smaller relative to a human, so easier to dodge; has less mass, so collides with less force (“bumps” is probably a better word); has lower ground clearance, so can’t run you over.

        I’m sure the AI is easier too. Imagine an AV car trying to make a judgment in an ambigous situation about whether it should speed up, slow down, or change lanes to avoid what it thinks might be a reckless lane-changing vehicle ahead, all while minimizing the chance of colliding with vehicles behind and beside. It must make a prediction, and it must make it now. It can’t stop and wait for the situation to resolve (well, it could – but that’s also likely to cause a collision). A confused sidewalk AV has the luxury of time. It can afford to stop and wait to figure things out. The worst case is that someone (probably on a phone) walks into it – or perhaps it gets hit by a reckless bicycle or scooter rider, but I think most people would say that’s on them.

        I don’t think AV cars can work in mixed urban traffic unless a) they become killing machines, b) they are so careful that they are far slower than willing-to-run-you-down human-driven cars, or c) we make the streets “safe” for them by tagging or removing the people. Possibility a) entails lawsuits. Possibility b) is no problem, but I think it would lead AV supporters to demand c). Result: either there’s no problem, or there’s a crisis point at which we can organize resistance. With sidewalk robots, I see people grudgingly giving them space until they become an established fact and it’s too late to take back our sidewalks. There is no crisis, no concerted resistance, and the outcome sucks. We need to ban them in advance.

        I think there’s a comparison with cell phone etiquette. We basically have none, and now it’s too late. The Japanese, in contrast, decided early on rules like banning phone conversations in most areas on their trains. I assume those norms have stuck. I think we need to deploy mainly laws against AVs, and mainly norms against scooters – now, before it’s too late.

        1. Since it is unacceptable that either a car or a cart hit anything or anybody their speed and the damage done is moot. Neither are ready for deployment until the risk is essentially zero. I would argue that even with all the bad driving out there that sidewalk behaviour is far far less predictable than road behaviour and the AI would be far more difficult for the cart on the sidewalk.

          Cars can travel as fast as they do because of the road being divided by direction, lane markings and curbs, traffic signals, turn signals and brake lights and generally the regulation of what may enter the roadway and when – none of which applies to sidewalks. (Not saying this is for the better – it just is.) This is the case even with “complete streets” – the road portion is far more regulated than the sidewalk. Where things are looser yet, Granville Island comes to mind, traffic generally drives very slowly for the very reason that nothing is clear and pedestrians routinely meander onto the road. AVs would similarly travel at very slow speeds and stop on a dime.

          Your point about “figuring things out” seems funny. Once the AI is developed enough to deploy AVs they will figure things out a lot faster than humans. And where humans are impatient and invincible AVs might actually slow down in “confusing” or congested situations or where road/weather conditions warrant.

          One could argue that AI might be ready to navigate autonomous cars on our roads once autonomous carts are capable of behaving appropriately on sidewalks. Not because it is easier, but because it is harder.

          But I completely agree that the conditions in which either type of vehicle would operate must be determined by us humans in advance and not by corporate interests. And their use of our roads or sidewalks should certainly not be free.

    2. Like hyperloop, hydrogen fuel cells, Segways or zeppelins these delivery robots are a good idea in principle, but will fail due to a very narrow segment where they could be commercially deployed. It will work ONLY in dense cities areas, say Yaletown or West End with its are many highrises and will be constrained to grocery or food delivery for the very few. For suburbs or larger areas it is too long a ride over too bumpy a terrain with many sidewalks and streets to cross. Prices of these machines will be high and maintenance costs astronomical. Sure, there may be the odd Davies Street restaurant trying one out for some initial publicity, but it will soon abandon its trial as too few customers use it and the cost per delivery is too high.

      It’s a toy. Good for marketing and hype. It’ll fade away shortly.

  3. I hear the ambulance chasers are already preparing the TV adverts: “Have you or a loved one been injured by an Autonomous Delivery Device? You may be entitled to compensation. Call the law offices of Ben Dover and C. Howitt Feils for more information…”

Leave a Reply

Your email address will not be published. Required fields are marked *