Autonomous Street Sweepers, aka Roomba for Public Works


I recently attended a Nashville WalknBike outreach event where the city presented their proposal for bike and sidewalk infrastructure development to meet the needs of our rapidly growing population. One of the key takeaways from the meeting is that there is a need for effective, inexpensive street cleaning.

The “state of the art” in street cleaning today for bike lanes are smaller and narrower versions of existing street cleaning vehicles. These vehicles typically cost in the $100-200k range and require one or two human operators. PeopleForBikes has done an excellent survey of the options available for public works departments. Some cities, such as Los Angeles, are testing smaller vehicles (Advance SW8000) which may be a better fit for tight bike lanes.

Sidewalks and bike lanes are an obvious and excellent application for autonomous, electric sweepers. Moving away from a human-crewed cabin vehicle gives you a long list of advantages:

Surprisingly a brief search does not turn up any existing commercial offerings of autonomous street sweepers. Design concepts exist – like the Scarab concept by Olga Kalugina – but so far remain thought exercises only.

It’s also worth noting that new capabilities do not necessarily eliminate jobs, but could be viewed as a force multiplier for existing service personnel. Depending on application, if the vehicles are not road-worthy it may make sense for humans to transport and deploy these bots during their cleaning cycle, allowing for more specialized bots without high-speed operations requirements. Additionally, some areas will simply be a better serviced by humans and existing sweeper equipment than by autonomous sweepers – the target is the majority of the street cleaning need that can be serviced by autonomous sweepers today, instead of more general-purpose and significantly more complicated cleaning robot.