Listen to this story
‘Scout’, a six-legged autonomous home-delivery robot by Amazon, started delivering packages in Snohomish County, Washington, for the first time in January 2019. Scout, like a good samaritan, used the sidewalks to travel. Upon reaching its destination, Scout would stop at the front door of its customer and open the lid so that the customer could collect their parcel. However, three years later, before Scout could fulfil its potential to be fully autonomous, Amazon just scrapped the whole project.
Amazon spokesperson Alisa Carroll said there were aspects of the programme that weren’t meeting customers’ needs.
Over the years, Amazon has invested heavily in robotics and is expanding its division with multiple acquisitions. Earlier in August, Amazon agreed to acquire iRobot for USD1.7 billion. Scout was also born after Amazon acquired robotics firm ‘Dispatch’ in 2017.
Reportedly, Amazon had around 400 of its employees working on the project. While Amazon claims that it is not shutting down the project entirely and rather reorienting the programme, it raises a much deeper question—Is autonomous tech doomed?
Why did Amazon decommission Scout?
After Amazon used Scout to deliver parcels in Snohomish County, it expanded to different locations across the US. By 2020, Amazon tested Scout in Atlanta, Georgia, Franklin and Tennessee. Back then, speculations were made that Scout could make delivery agents redundant. So, what went wrong with Scout?
Several reports suggested that Amazon is closing down its experimental projects as gross sales progress slows after the highs from the pandemic. During the second quarter of 2022, Amazon reported a growth rate of 7% year-on-year; however, it was the e-commerce giant’s slowest growth in over two decades.
Earlier this month, Amazon also discontinued ‘Amazon Glow’—its interactive, video-calling device for kids and families—just one year after its debut.
Interestingly, both ‘Scout’ and ‘Amazon Glow’ were products that emerged amid the pandemic. Scout was Amazon’s solution to ensure contactless delivery, whereas Glow was released keeping in mind that everyone was isolated at home due to the lockdown measures introduced across the globe. However, now, with the world getting back to normalcy, products such as ‘Glow’ were not able to find many takers.
For people at Amazon, Scout might have seemed like a good idea amid the pandemic; however, how feasible are such robots? Hussein Kanji, an early-stage venture capitalist, in fact, called the Scout project a ‘dumb idea’.
Paris Marx, who is the host of the podcast ‘Tech Won’t Save Us’ said, “Like the delivery drones, the autonomous robots to replace delivery workers never made much sense. They’d need the sidewalks to turn into their roads, and even then, the distances are still a problem. Hopefully, we will see more of them shut down soon.”
Autonomous vehicles: Still a dream?
Besides financial stress, another aspect that could have forced Amazon to shut down its autonomous home delivery project is that they don’t have the bandwidth at all to crack the autonomous technology.
The dream of autonomous vehicles is not really new, and with over USD100 billion spent so far, the progress made is not at all satisfying. To be precise, the optimism surrounding autonomous vehicles has been fading for some time now.
Anthony Levandowski, often regarded as the godfather of autonomous vehicles, said in an interview with Bloomberg, “You’d be hard-pressed to find another industry that’s invested so many dollars in R&D, and that has delivered so little.”
Tesla CEO Elon Musk has also been promising driverless Tesla cars or Robotaxis for almost a decade now. But, Tesla cars have been plagued by numerous accidents—accounting for nearly 70% of reported crashes.
“The whole autonomous car/robot complex is crashing down just like I predicted years ago it would. The reason is simple: we have not figured out Moravec’s paradox, and our statistical-in-nature approach to AI is unable to deal with data generated by complex systems,” Filip Piekniewski, software engineer/scientist at Accel Robotics, said.
Scout, which was meant to be an autonomous home-delivery robot, was often accompanied by a human, like a father taking their child door-to-door during Halloween. There is a possibility that Amazon realised a fully autonomous Scout was never going to happen, at least in the foreseeable future. This realisation could very well be the reason Amazon has not completely abandoned Scout but, in fact, is reorienting the programme.
It would be interesting to see what the future holds for Zoox-Amazon’s autonomous vehicle subsidiary. Amazon acquired Zoox for more than USD1 billion in 2020.