Ask HN: Robotics projects in consort with the City and build advice | Hacker News Reader

Ask HN: Robotics projects in consort with the City and build advice

2 pointsitissid9 comments

I am looking to build a limited mobility human in loop robot that augments and enhances an existing city function.

The robot is on 2–4 wheels and its environment is the city-street(not highway). It need not move too fast (< 4-10 km/h), and its environment is quite observable(via sensors and augmentation). It is under the supervision of a person all the time. Assume location is North America.

Here are some questions I don't know the answer to:

1. What kind of partnership works best in such cases, VC University-City collaboration? Assume the time horizon is 1-2 years for such a project

2. Do accelerators/VCs to mentor one through such a thing exist, and do you know of any?

3. What kind of regulations and legal code can kill a project like this(especially if it's like a for-profit startup) and how can I find it quickly within the city bureaucracy?

4. Have you built this kind of project or navigated a city bureaucracy to complete a tech project? If so, mind sharing your advice?

City street but going slow sounds like a bike-lane type robot. The most comparable existing robotics companies to that are probably Refraction AI, Vayu, and Intermode. Nuro is small, but still fast enough to be considered comparable to a car. So I would look at the former three and see how they are navigating this regulatory space (no pun intended). And no, most VCs don't mentor you through that, because most are not interested in companies that sell to cities, because it's a hard market to sell to, so they don't have any experience in it.

I was part of a project where we developed a Mobility Marketplace for some major cities. What I learned during that time is this: city politics is all about talking to everyone and their mother. Eventually the mayor needs to agree to it, and all he or she will do is ask everyone around them what they think. So you want everyone to give their blessing. "Everyone" here means all sorts of organizations from the department of transportation to business improvements consortia, to citizen groups, to the local AAA, the police, shop owner associations, you name it. You probably want to hire a consultant with that kind of network for intros and to coach you on how to approach each conversation. We had someone who had been selling to cities for over 20 years at the time and it was clear that we would have made many mistakes without her.

What's the application/city-function? Also, which city/state do you want to start in? It goes without saying that some cities/mayors are much more innovative than others.

Thanks for the advice! I am trying to augment street cleaning and change the alternate side parking to make it better. It happens throughout the big cities in the dense urban parts like NYC, Jersey City, SF, LA.

Mine is a small city in hudson county NJ.

I've been talking to a few people and here is what they had to say to add onto what you already mentioned:

1. Ask the city if they already have some future investment in the area, that will help me narrow down what department and people to talk to.

2. Look at the city's budget on street cleaning. If it's say 5 million a year and there is some revenue generation say 2.5 million, then how does what I am doing affect it.

Once I have these narrowed down these and see no obvious red flags to govt approval, I can build a simple landing page and survey to gauge public interest.

Do you know https://www.viabot.com/ and https://www.enway.ai/ ?

Thanks. I hadn't heard of them. But it would be really cool to connect with them and learn about some crossover learnings I am sure I will encounter down the line.

What would you say your competitive advantage is compared to these companies that have a few years of head start? Don't mean to discourage you, but it's better to think about this early.

I think these companies are trying to solve a different problem: They are just inventing a way to clean streets using near autonomy.

I am trying to figure out how to solve a pain-point which can work along with current or these technologies and I can bring them faster to market because they are not nearly as complicated (human supervision). Without elaborating the idea completely here(being cautious), it solves a problem unique to us that are not being solved by any one of these companies.

Interesting. I have more questions but this thread is getting long. Want to find me online (see my profile) so we can chat elsewhere?

Investors won't be interested in a one-shot project. They'd be interested if the company can grow big, either by selling systems to other cities, or expanding into related businesses. So figure out whether or not that's possible.

Most cities are terrible customers. They make decisions slowly and often impose painful requirements on suppliers. They're wary of anything that could fail publicly, because the media loves to show failures and blame it on the city government.

Another way that cities are bad customers is by not actually using things they buy. They might buy your product and then not get around to hiring anyone to operate it. Even though you got paid, you won't be able to iterate and improve the product.

City purchases are generally public records, so you should look for similar products the city has bought and go talk to people in the company that made it for advice and connections.

Thanks a ton. That is great advice for strategy!

Reply on news.ycombinator.com