Artificial Intelligence is Reshaping Public Transit as We Know It

00:00 Levi McCollum: Stop Requested. Welcome to Stop Requested, the podcast where we discuss everything transit.

00:05 Levi McCollum: I’m your co host Levi McCollum, director of operations at ETA Transit.

00:10 Christian Londono: And I’m your co host Christian Londono, Senior Customer Success Manager at ETA Transit.

00:20 Levi McCollum: Hey Christian, how’s it going this afternoon?

00:22 Christian Londono: Hey Levi, how’s it going?

00:24 Levi McCollum: It’s going pretty well. So what are we talking about today?

00:28 Christian Londono: You know Levi, the topic today I think is quite exciting.

00:32 Christian Londono: I think you’ve used ChatGPT. I think everybody has at this point that is listening to our episode and knows

00:40 Christian Londono: about this technology which is powered as generally we know by AI artificial intelligence.

00:48 Christian Londono: So that is our topic today.

00:50 Christian Londono: Our topic is artificial intelligence and machine learning and how these technologies are being implement it in public transit and

01:00 Christian Londono: have the potential to impact the future of public transportation.

01:04 Christian Londono: So I think is a. Is a very interesting topic.

01:07 Levi McCollum: Absolutely. Now this is exciting. Of course the technology in and of itself is exciting, but when you apply it

01:13 Levi McCollum: to public transportation, something that we are deeply familiar with and we, we love so much, I think this is

01:20 Levi McCollum: really where the nerd in me comes out because we get to talk about all the what could be, you

01:25 Levi McCollum: know, the what ifs when you apply AI and machine learning to public transit.

01:31 Levi McCollum: So I’m looking forward to this.

01:32 Levi McCollum: It’s going to be good.

01:33 Christian Londono: Oh, it’s going to be fun.

01:34 Christian Londono: And technology tends to revolutionize the way that we do things and take us to more efficiency, better service.

01:43 Christian Londono: And we’ve seen it with different technologies.

01:45 Christian Londono: And I personally believe that AI and what we’re seeing today, it’s another industrial revolution and it’s definitely going to

01:55 Christian Londono: impact the public transportation industry. And I think to get started into this conversation, as we were preparing for today’s

02:04 Christian Londono: episode and doing some research, I personally could not really see the difference between artificial intelligence and machine learning or

02:13 Christian Londono: ML. And I think that for all of our listeners that will be a great place to start, which is

02:19 Christian Londono: having an understanding at least if in a simple way.

02:22 Christian Londono: G give us something simple about this.

02:26 Christian Londono: So everybody can start with understanding generally.

02:29 Christian Londono: What’s artificial intelligence in machine learning?

02:31 Levi McCollum: Yeah, I mean that’s a good question and I think that’s one that we have to set from the very

02:36 Levi McCollum: beginning here. But I think the idea behind artificial intelligence is the concept of a machine learning to behave or

02:47 Levi McCollum: act or sense the world like a human would.

02:50 Levi McCollum: And machine learning would be the machine’s ability to be able to extract that knowledge and then work autonomously from

03:00 Levi McCollum: that knowledge. Probably the easiest way to look at it is in terms of artificial Intelligence being the umbrella term

03:11 Levi McCollum: that people would use for all different types of these intelligences, like robotics.

03:17 Levi McCollum: Right. You might be familiar with Boston Dynamics, who create these humanoid type robots.

03:23 Levi McCollum: Those robotics fall under AI. Right.

03:26 Levi McCollum: Machine learning or large language models also fall under AI.

03:31 Levi McCollum: So we’re familiar with ChatGPT, but that’s a large language processing model.

03:36 Levi McCollum: And again just kind of all sits under the umbrella term, this larger concept of AI.

03:41 Levi McCollum: So I’m not an expert in it, but hopefully that clears it up or maybe it made it more difficult

03:46 Levi McCollum: to understand. I’m not sure. We’ll see.

03:48 Christian Londono: I think it was a good, simple definition.

03:53 Christian Londono: And more important, and as is relevant to this episode is why do you think that’s important for public transit?

04:01 Christian Londono: Why would people in public transit be thinking about artificial intelligence?

04:07 Levi McCollum: Yeah, I think it’s to improve the service.

04:10 Levi McCollum: Right. That’s what we’re after. It’s a public good, it’s a public service.

04:14 Levi McCollum: And using artificial intelligence or machine learning, either of these applications here could bring about a better service that benefits

04:28 Levi McCollum: the customers lives. I think that’s really what we’re after.

04:32 Levi McCollum: I know we’re going to talk about some examples and look at some case studies and so on, but you

04:37 Levi McCollum: want to make sure that the service is operating well, that it’s not only efficient, but that it’s safe.

04:44 Levi McCollum: And I think as we’re going to get into it, AI can really help on the safety side, maybe more

04:50 Levi McCollum: so than other areas. Of course, we’ll see after the end of the conversation today.

04:55 Levi McCollum: But yeah, making it better.

04:58 Christian Londono: Yeah, I think that it has the potential to optimize operations and enhance user experience, like you said.

05:05 Christian Londono: And it’s a lot of automation, it’s automating a lot of the things that we do today where we’re bringing

05:12 Christian Londono: the power of a computer, a human being is able to do a lot of things that you bring, the

05:20 Christian Londono: automation, you bring the intelligence of a computer.

05:23 Christian Londono: And now it can enhance, of course, the efficiency and just the consistency as well.

05:30 Christian Londono: And I think that when we talk about current applications and what we’ve heard, we’ve seen come up in publications

05:42 Christian Londono: and magazines about transit agencies being innovative, some of those agencies that have an innovation department and they’re dabbing into

05:53 Christian Londono: technologies and they’re trying to see how those can be brought to public transit.

06:00 Christian Londono: And you know, I really have to say that I’m really appreciative of those agencies.

06:04 Christian Londono: I know in Florida we have jta, Jacksonville Transit Authority, and it’s one of many that are very innovative, but

06:13 Christian Londono: they’re one that particularly they stand out and they have a team that are dedicated for innovation in particularly we’ll,

06:23 Christian Londono: we’ll talk about some future technologies that use AI, but they’re already looking into it and, and they do a

06:31 Christian Londono: great service for the industry. Right.

06:33 Christian Londono: Because it takes the first one and a lot of times nobody wants to be the first one.

06:39 Christian Londono: It’s risk in trying to think there’s a risk.

06:42 Christian Londono: And you have to take measured risk, you know, calculated risk and you have to reiterate and then, you know,

06:50 Christian Londono: maybe expand a little. You of like start small and start expanding.

06:54 Christian Londono: But definitely those agencies pathway for all these other agencies to be able to just copy and paste and then

07:01 Christian Londono: benefit from, from the same, you know, enhancement or, or technology.

07:06 Christian Londono: But one of the technologies in AI that I wanted to kind of like tell you about today and bring

07:12 Christian Londono: to the discussion is about predictive maintenance.

07:16 Levi McCollum: So, so what does that mean?

07:19 Christian Londono: So generally, I mean, you heard about preventive maintenance.

07:23 Levi McCollum: Yeah, preventative maintenance. Right.

07:25 Christian Londono: So preventative maintenance is kind of like a good practice.

07:29 Christian Londono: I mean it’s a requirement in terms of our, you know, safety.

07:33 Christian Londono: You know, you talk about a ptasp, you know, your maintenance plan is included, which includes a lot of components

07:44 Christian Londono: of preventive and preventative maintenance.

07:47 Levi McCollum: Right.

07:47 Christian Londono: You have to do every so many miles to the vehicle to get the vehicles in good working conditions, avoid

07:55 Christian Londono: roll calls, breakdowns, and potentially a safety hazard.

07:58 Christian Londono: Right. Like if something was to go bad in a vehicle that might cost a vehicle to get into an

08:03 Christian Londono: accident, that could be, you know, a big tragedy.

08:07 Christian Londono: Right.

08:07 Levi McCollum: So the idea is if you do those periodically over time, then you’re going to reduce the number of accidents

08:13 Levi McCollum: and incidents that you have to respond to.

08:15 Christian Londono: Right. And that’s been the practice in terms of preventative.

08:20 Christian Londono: Right. But then when you talk about AI, now we’re starting to hear about predictive maintenance.

08:26 Christian Londono: And the first time I heard about it, I was kind of excited.

08:29 Christian Londono: I was like that. That seems kind of cool because it’s kind of like reading the future.

08:35 Christian Londono: It’s like somebody saying that vehicle is going to break down tomorrow.

08:39 Christian Londono: So how powerful is that information? How powerful would that information be for you today? Yeah, so I think that

08:47 Christian Londono: application is really interesting. I know some agencies in Florida have started working with vendors and using all of their

08:56 Christian Londono: historical data. I honestly want to look forward to learn more about it and to hear about the benefits.

09:03 Christian Londono: I know high level, from what I’ve heard, it’s very promising and you know, they speak very highly of it.

09:10 Christian Londono: Just think about not just kind of like following the same a, you know, preventative maintenance, which is like the

09:20 Christian Londono: same things you do every so many miles.

09:22 Levi McCollum: Right.

09:22 Christian Londono: But then using all these data that is based on your historical breakdowns and, and you know, what happened to

09:30 Christian Londono: start telling you what’s going to happen for you and then get ahead of that.

09:33 Christian Londono: And that’s the game. The game in terms of, of maintenance is reducing mechanical breakdowns.

09:39 Christian Londono: Like the vehicles are not failing because of something that maybe we couldn’t address on the vehicles before the vehicle

09:46 Christian Londono: went out of service. So that, that translated to a better experience for the users not being impacted by a

09:52 Christian Londono: mechanical breakdown. But that’s, that’s definitely one that I think is, is today is having a great impact in the

09:58 Christian Londono: transit industry.

10:00 Levi McCollum: Yeah.

10:00 Christian Londono: And I’m just, you know, curious.

10:02 Christian Londono: What’s another one that you heard about that you maybe want to bring up?

10:06 Levi McCollum: Yeah, well, I, I think that your example is a really solid one.

10:09 Levi McCollum: I like that a lot. And maybe this also ties into my example that I’m about to give as well.

10:15 Levi McCollum: But there’s so many data points that the bus is giving off.

10:19 Levi McCollum: I think we talked about this in an episode or two back.

10:21 Levi McCollum: It’s like the bus is this big piece of high tech now.

10:25 Levi McCollum: It has so much data that it’s throwing off that one person is not able to really make sense of

10:32 Levi McCollum: that. Right. A human brain might be able to look at bits and pieces, but you know, to conceptualize these

10:41 Levi McCollum: thousands or hundreds of thousands of data points and then make some sort of action out of that nearly impossible

10:48 Levi McCollum: for a human brain, it’s super time consuming.

10:50 Levi McCollum: You would need someone with advanced statistics and, you know, understanding mathematics, understanding.

10:56 Levi McCollum: And that person has to be present at a transit agency.

11:00 Levi McCollum: There are just so many complications there, I think, with making sure that that happens.

11:05 Levi McCollum: But you know, to, to that point, planning and scheduling, I think is one area where we’re starting to see

11:12 Levi McCollum: the AI terminology start to creep in.

11:16 Levi McCollum: Of course, you know, there are a lot of data points there.

11:19 Levi McCollum: The CAT avl, like our ETA CAD avl, for example, you know, it’s sucking in all this data from the

11:25 Levi McCollum: bus. It’s saying, okay, this is how many miles you travel, this is how fast you were going, this is

11:29 Levi McCollum: who’s driving the vehicle. This is when they stopped and this is when they started going again after they did

11:36 Levi McCollum: layover. We know exactly which stops you were at.

11:41 Levi McCollum: You put all that together over hundreds of thousands of hours for a particular driver and you get a pretty

11:50 Levi McCollum: good profile of how that driver operates a bus, what’s their Average speed, which vehicle do they like, which route

11:58 Levi McCollum: are they on based on their, their assignment for that particular period? So, you know, you can start to get

12:05 Levi McCollum: a profile of an operator. You can adjust your schedules accordingly based on how they’re actually operated.

12:12 Levi McCollum: I mean, I wish so many times that I had something that was a little bit more intelligent than me

12:18 Levi McCollum: telling me, or a lot more intelligent than me telling me, like, hey, this is what’s wrong with the schedule,

12:24 Levi McCollum: right? It’s this part right here.

12:25 Levi McCollum: It’s this segment that, you know, you scheduled 15 minutes, but really you need 20.

12:31 Christian Londono: And it’s the highlighting of what you need to pay attention to.

12:35 Christian Londono: Right? Because you know, just listen to what you said about transit planning and it brings me back to my

12:43 Christian Londono: days and is also there’s a lot of data analytics that you do when you’re a transit planner.

12:49 Christian Londono: But it’s also knowing what data to look at, like you said, is you will have to spend tremendous amount

12:55 Christian Londono: of hours running so many reports and looking at all these data points to start to get the data insight.

13:00 Christian Londono: And you know, I’ve understood from or I heard from people that work in this space in terms of data

13:06 Christian Londono: analysis and all that, and it’s like 80% of the time is data mining and 20% is the data analytics

13:14 Christian Londono: where you actually do the data inference and extract the insights of the data.

13:18 Christian Londono: So what you’re talking about is with AI, we’re reducing that data mining and trying to find those sets of

13:29 Christian Londono: information that we need to take a look at and that is also actionable.

13:32 Christian Londono: What is this telling me? How can I use this to make service changes?

13:37 Levi McCollum: Right? Because it’s one thing to have the data, but if you’re not taking action from the data, you might

13:43 Levi McCollum: as well not even have it.

13:44 Levi McCollum: Honestly, it does zero good to just be collecting it and digitally it’s collecting dust, it’s just there.

13:52 Levi McCollum: But you’ve got to be able to make use of it.

13:55 Levi McCollum: And if AI is that potential, you know, disruptor, that’s going to allow agencies some more freedom with understanding the

14:04 Levi McCollum: data some more time back to be able to take care of other tasks and to run a more efficient

14:08 Levi McCollum: service. I mean, heck yeah, I’m sign me up for that, I’d love it.

14:12 Christian Londono: I’m known for that too. And the thing is, I think many agencies, and we’ve been guilty of that, collect

14:19 Christian Londono: a bunch of spreadsheets and there’s so many reports in different places.

14:24 Christian Londono: And it’s not just the collecting of the data.

14:26 Christian Londono: I think there’s a lot of importance to that as well.

14:29 Christian Londono: But it’s really the data insights and what are you doing about it and being help with that process.

14:36 Christian Londono: Yeah, that’s definitely a great application.

14:38 Christian Londono: Another thing that comes to my mind is the overall passenger experience and how it can be enhanced with things

14:45 Christian Londono: like chatbots and visiting some websites this last week of different customers, transit agencies out there, I keep seeing more

15:01 Christian Londono: often customer service chatbot popping up.

15:06 Christian Londono: So as you’re going to the website and I think it doesn’t take that many seconds that you’re in that

15:11 Christian Londono: landing page and then a pop up comes up which is a bot asking you if they can help you

15:17 Christian Londono: with anything. What are you trying to find? How can I help you navigate? And it’s very interesting because you

15:22 Christian Londono: know, in a previous role that I had, you know, overseeing also the website for a transit agency and being

15:30 Christian Londono: part of the discussion how can make it, you know, we can make it better.

15:34 Christian Londono: We’re talking about the navigation of the site and then how easy is to navigate and to find the things

15:40 Christian Londono: and then using the role of a customer and different people that will be coming to visit your side and

15:46 Christian Londono: finding stuff. But the bot, it cuts a lot of that.

15:49 Christian Londono: I mean it’s, it pretty much is asking you how can I help you? What do you need?

15:52 Levi McCollum: Right.

15:53 Christian Londono: And I think it’s really powerful not only in helping you find the information but also it also has a

16:02 Christian Londono: lot of applications. So sometimes you can load up predetermined answers or have a lot of information of the agency

16:11 Christian Londono: that it can quickly provide to that customer, to that visitor.

16:14 Christian Londono: And we’re talking about just the website but there’s other outlets with something similar is being implemented.

16:20 Christian Londono: It could reduce calls time that you know, you’re using customer service agents to try to answer some of these

16:27 Christian Londono: questions. But then some agencies are doing a combination of that too where if they matter, your, your, your matter

16:37 Christian Londono: that you’re trying to get resolved or the answer that you’re looking for is quite complex.

16:41 Christian Londono: It can actually connect that person with a life agent that is going to be answering those questions.

16:47 Christian Londono: So I, I think that’s, that’s a good application at least for that pie websites.

16:52 Levi McCollum: Yeah, I mean a chatbot may not be able to solve all the problems right now and I understand that

16:59 Levi McCollum: that’s where the chatbot would pass it along to that customer service agent.

17:03 Levi McCollum: So you know, you have a real life human there that’s also able to assist and see the chat history

17:08 Levi McCollum: I would presume and be able to direct the individual based on the request from there.

17:14 Levi McCollum: Is that how that would work?

17:15 Christian Londono: Oh, yeah. And you know, customer service calls.

17:20 Levi McCollum: And you were in customer service.

17:22 Levi McCollum: No. Let’s hear it.

17:23 Christian Londono: I, I’ve done customer service. I’ve been a customer service agent.

17:26 Christian Londono: I had the headset on and, and answer calls from live agents.

17:30 Christian Londono: I mean, for, for live writers, customers.

17:33 Christian Londono: And I’ll tell you one thing, you know, a great deal of the calls are about travel suggestions or travel

17:41 Christian Londono: indications. This is where I am.

17:43 Christian Londono: This is where I need to get.

17:45 Christian Londono: You know, sometimes, of course, a lot of codes are terms.

17:48 Christian Londono: ETAs is the next boss. You know, I’m at this location, where’s the bus.

17:52 Christian Londono: Right. But I get a great chunk is how do I go places? And AI is now being able to

18:00 Christian Londono: give kind of like personalized travel suggestions and, you know, helping people choose, you know, how they’re going to travel

18:06 Christian Londono: and find this information. So that’s another, you know, a case of, of reducing calls, reducing people calling your customer

18:15 Christian Londono: service agents, and then having somebody manually try to find information.

18:19 Christian Londono: What’s your origin, what’s your dest.

18:21 Christian Londono: Destination, and kind of like pull this information.

18:24 Christian Londono: Have you tried AI for personalized travel?

18:28 Levi McCollum: Yeah, so I did try it.

18:30 Levi McCollum: Now, I will put this caveat out there that it was at least a year ago, so fairly early on

18:36 Levi McCollum: in the whole chatgpt and AI craze.

18:40 Levi McCollum: But I wanted to see how well it would plan my trip and here in West Palm Beach.

18:45 Levi McCollum: And I don’t remember exactly the, the origin and destination that I put in, but, you know, something to the

18:52 Levi McCollum: effect of going from West Palm beach to Boca Raton.

18:55 Levi McCollum: And it did not give me very good results.

18:59 Levi McCollum: It made up routes, it made up times that didn’t align with the schedule.

19:03 Levi McCollum: It said I could transfer places that just didn’t exist.

19:06 Levi McCollum: So there were a lot of hallucinations then.

19:09 Levi McCollum: I would think that it’s better now.

19:11 Levi McCollum: However, just like the, the disclaimer that Chat, GPT and others, you know, other providers have at their the bot

19:20 Levi McCollum: page where it says you got to check the details.

19:22 Levi McCollum: Right. I mean, you still have to do a little bit of work.

19:25 Levi McCollum: It’s not going to do everything regarding your trip planning for you.

19:29 Levi McCollum: But now, how cool would that be, though, if the experience were that seamless and if the results were right?

19:37 Christian Londono: Yeah. You know, as you were saying, that something came to my mind that is kind of scary.

19:44 Christian Londono: So I think that the. For your case, particularly especially Palm beach county, and you serve as the director transit

19:53 Christian Londono: planning for Palm Trend, which means that you were buried in the schedules on the Service, you understood the service

20:00 Christian Londono: very well. So I think that that 100 help to spot that.

20:04 Christian Londono: The answer from ChatGPT is like, no, this doesn’t make sense.

20:09 Christian Londono: What I’m thinking is agencies definitely want to look into it.

20:14 Christian Londono: Everybody that is listening, I would invite you to try ChatGPT or Copilot or any of the AIs that is

20:25 Christian Londono: out there to see if gives you answers in terms of trip planning and how accurate is it, if it’s

20:31 Christian Londono: working. Because I’m wondering if your customers today, especially if, let’s say that you’re in a college town and you

20:38 Christian Londono: have a lot of young kids, that they’re way ahead of any of us in terms of using these technologies

20:44 Christian Londono: and they might be asking for their travel plans, starting to dab to or, or change from calling customer service

20:51 Christian Londono: and your transit agency and then starting to do AI.

20:55 Christian Londono: And then maybe they’re following some of these, you know, indications for using public transit and maybe they’re wrong.

21:04 Christian Londono: Right, so, so I think it will be important and I would invite listeners to give it a shot and

21:10 Christian Londono: see if it’s giving something good.

21:12 Christian Londono: Because if it is, let’s say that they’re testing it and I think Liba, you and I, we might want

21:16 Christian Londono: to test it again to see what comes out now a year after.

21:20 Christian Londono: But if it’s giving good information, then that’s something that you might be able to kind of capitalize, right?

21:30 Levi McCollum: Yeah, absolutely. I think that if it’s working and again, that was a year ago, that this technology is changing

21:37 Levi McCollum: by the day, it seems with new models coming out from even new competitors.

21:42 Levi McCollum: Right, like brand new competitors that have, have kind of sprouted up over the last six, eight months.

21:49 Levi McCollum: But yeah, there are so many models out there that are competing.

21:52 Levi McCollum: Of course you’ve got Google’s Gemini and then you have Perplexity.

21:56 Levi McCollum: But if you’re able to test across multiple models and able to get somewhat reliable results, I really think that

22:06 Levi McCollum: this does have, I hate to just throw around this term, but revolutionary, right? I mean, that is pretty revolutionary

22:13 Levi McCollum: at least in terms of our space, in terms of the, you know, public transit world, you know, is it

22:19 Levi McCollum: autonomous vehicles level revolutionary? No, but it has its own place.

22:24 Levi McCollum: I think on the rung of being pretty revolutionary.

22:26 Levi McCollum: If, if my trip planning can now be, you know, broken down into this large language model type, type based

22:33 Levi McCollum: format. I think that, you know, it frees a lot of people up to be able to do other things

22:39 Levi McCollum: rather than, you know, be on the phone saying, yes, if you get on this bus at this time and

22:43 Levi McCollum: then transfer here, then you can get to this other place, you know.

22:48 Christian Londono: Yeah.

22:49 Levi McCollum: I just think that there are some ways for us to be able to improve public transit and certainly that

22:54 Levi McCollum: customer facing side of it is one area that I think AI could help in.

23:00 Christian Londono: Yeah, I think in our current climate today, where we have to be very conscious about our expenses in terms

23:09 Christian Londono: of public transit and be efficient in the way that we deliver it, anything that could enhance our operational efficiency,

23:20 Christian Londono: it has the potential of yield savings even if they’re small.

23:24 Christian Londono: You know, like maybe, you know, you could reduce the customer service staff and then you could use those resources

23:30 Christian Londono: somewhere else. Right. I, I would take it.

23:33 Christian Londono: And definitely something that all transit agencies should be paying a close attention to.

23:40 Christian Londono: And in that vein, what other area it’s being currently impacted by AI where we’re seeing applications and it’s helping

23:56 Christian Londono: agencies today?

23:58 Levi McCollum: Yeah, one other area I would call out is safety.

24:03 Levi McCollum: Safety. As I said from the top of the podcast, this is one area that could be, be pretty significantly

24:10 Levi McCollum: improved by AI. Yeah. And you know, you want to make sure that your passengers are safe in and around

24:17 Levi McCollum: the transit facility on board the transit vehicle.

24:22 Levi McCollum: So it, it does matter that AI is being applied in a, in a way that is going to benefit

24:29 Levi McCollum: the most people. Right. And our day and age, it is, it is on a transit vehicle where some unfortunate

24:38 Levi McCollum: and bad things can happen. People can bring weapons on board a transit bus or a train perhaps, or into

24:47 Levi McCollum: a train station. You know, these are the worst case scenarios that we’ve seen play out across the world unfortunately

24:54 Levi McCollum: way too many times. You know, there are some technologies now that will recognize weapons when they’re brought into a

25:04 Levi McCollum: train station or when they’re brought onto a, a bus.

25:06 Levi McCollum: If you have, you know, a gun on you, perhaps that can be detected and the operator of the vehicle

25:13 Levi McCollum: can be alerted, or if you have some security staff or if you have a police department that’s a part

25:20 Levi McCollum: of your transit agency, then they could be alerted and respond accordingly.

25:24 Christian Londono: Yeah, those are, those are very important, especially the weapons detection.

25:30 Christian Londono: And you know, I’ve heard about agencies kind of using this application with the cameras where the intelligence behind it

25:37 Christian Londono: can, you know, do the part where is identifying weapons and bring up alerts.

25:44 Christian Londono: I’ve also heard about packages, so abandoned packages, fortunately public transit, especially in large systems, in transit hubs or train

25:57 Christian Londono: stations, those are targeted a lot of the times for some of these atrocious things that happen in transit, different

26:05 Christian Londono: places around the world. So I’ve heard about using that technology that surveillance camera system where it can identify packages

26:12 Christian Londono: almost like weapons and, and also send alerts.

26:16 Christian Londono: I’ve also heard about using the camera system and AI for variation where it’s able to recognize when people are

26:26 Christian Londono: jumping over, you know, the turnstile, turnstiles, and, you know, they’re not actually paying or validating before they can actually

26:35 Christian Londono: be granted permission to. To actually go through.

26:39 Christian Londono: And that’s actually led to transit agencies changing their fare gates and now putting more of a physical barrier to

26:47 Christian Londono: try to, you know, minimize the variation.

26:50 Christian Londono: I know that was a big topic last year.

26:54 Christian Londono: So, yes, all together, the safety is definitely paramount in transit.

27:00 Christian Londono: And there’s different AI applications we’re talking about now.

27:05 Christian Londono: I think not just applications that are coming up, but also new applications that will be in the future of

27:14 Christian Londono: public transit. And keeping in that vein of safety, I know one of the big technologies that we’ve heard about

27:23 Christian Londono: is autonomous vehicles and how it’s being implemented into public transit and the potential.

27:30 Christian Londono: You know, here we talk about, in artificial intelligence, there’s a lot about eliminating sometimes the human factor, which humans

27:40 Christian Londono: tend to make more mistakes than machines.

27:42 Christian Londono: Right. And then when we talk about autonomous vehicles is putting a computer to drive the vehicle.

27:48 Christian Londono: Right. And potentially that could yield reduction on safety scenarios.

27:54 Christian Londono: Right. So that’s definitely part of what we see in the future for public transit.

28:00 Christian Londono: We know of different transit agencies doing small autonomous vehicles, kind of like loopers in some downtowns, and they’re starting

28:09 Christian Londono: to adapt into this technology. Anything else you heard, AI related into future, that is in the future public transit?

28:19 Levi McCollum: Well, maybe one that I should have mentioned earlier, it kind of falls into the realm of safety.

28:24 Levi McCollum: So I’ll go ahead and just throw it out there anyway, you know, being able to detect that a vehicle

28:30 Levi McCollum: is parked in a BRT lane, for example, a dedicated lane that’s being blocked by a vehicle that’s just illegally

28:39 Levi McCollum: parked. Right. That is a technology that is now in use, and it seems like advancing rather quickly.

28:46 Levi McCollum: And I know that there are several cities around the country, New York and la, at least those two, perhaps

28:53 Levi McCollum: even more now, that are utilizing that technology because they have a good number of dedicated lane miles.

29:01 Levi McCollum: So when you kind of add up the safety concerns and the inability to be able to make a schedule

29:12 Levi McCollum: based on someone that’s illegally parked in the lane, then that adds up.

29:19 Levi McCollum: Right. It impacts the performance of the vehicle on the route, and it really is a safety hazard.

29:26 Levi McCollum: Right. If someone were to pop out of a, you know, a driver’s side door and, you know, they run

29:33 Levi McCollum: into A bus. I mean, that, that would be worst case scenario and it’s a black mark on the agency.

29:39 Levi McCollum: You know, of course that person’s live there in that scenario.

29:43 Levi McCollum: That’s, that’s not good. So, you know, we, I think we’ve, we’ve got to look at those, those applications that

29:50 Levi McCollum: are, they’re nascent, but they are, they’re still applicable to, you know, to our world today.

30:01 Levi McCollum: It’s not thinking so outside the box and so far out into the future that is this really close to

30:09 Levi McCollum: being possible. Well, I mean, for that particular case of, you know, seeing the vehicles that are blocking a, a

30:17 Levi McCollum: dedicated lane. Yeah. I mean that, that is, that’s right here.

30:20 Levi McCollum: That’s here today. And you know, fortunately we have some people that are, you know, developing these technologies and improving

30:26 Levi McCollum: them over time.

30:28 Christian Londono: Yeah. And, and that’s changing that situation.

30:31 Christian Londono: Right. It has the potential of changing the, the, the reality.

30:35 Christian Londono: Right. I think that case you alluded to, if I’m not mistaken, that that’s New Jersey.

30:43 Christian Londono: Yeah, yeah. Where they were recognizing those vehicles.

30:46 Christian Londono: And I think anybody that was blocking the, you know, bus lanes and you know, I don’t know what the

30:54 Christian Londono: numbers are today, but I’m hoping that it reduce the amount of vehicles blocking the lanes, reducing potential hazardous situations

31:02 Christian Londono: and allowing the bus to make the schedule.

31:05 Christian Londono: Like you said, for the riders experience is supposed to be bus rapid transit.

31:10 Christian Londono: And then it’s supposed to be, you know, driving your car because, you know, you’re in mixed traffic.

31:16 Christian Londono: But if all these people. So in that blocking the bus, then is deterring people from actually benefiting from that

31:24 Christian Londono: purpose. I really like that.

31:27 Levi McCollum: Yeah. The dedicated lane doesn’t really matter then.

31:29 Levi McCollum: Right. You’re just as slow as all the other cars around you because you’re in the mixed traffic.

31:35 Levi McCollum: You’re right there alongside the soccer mom and the guy in the big lifted truck.

31:44 Levi McCollum: It doesn’t really do you any good.

31:46 Christian Londono: Oh, yeah, no. People have the audacity and, and you see, they’re trying to benefit.

31:50 Christian Londono: It’s funny that, because they’re trying to benefit of bus, you know, rapid transit, which is, you know, having dedicated

31:57 Christian Londono: lane. But in order to get a benefit, you have to get on the bus.

32:01 Christian Londono: Otherwise everybody will be in that lane and then nobody’s moving fast, everybody’s stuck.

32:06 Christian Londono: So, you know, it’s, it’s interesting, people just, you know, the best solution for that is, is just jump on

32:12 Christian Londono: the bus and then benefit from moving quicker than regular traffic.

32:17 Levi McCollum: Now some of these companies though, they will say, especially the ones that are Focused on evasion or the rule

32:25 Levi McCollum: breaking, let’s say like that example of a vehicle being illegally parked in a dedicated lane or the fare evasion

32:33 Levi McCollum: type technologies, they often say that, that they’re not in it to ticket people.

32:39 Levi McCollum: Like that’s, that’s not the objective is to ticket people, but it’s to understand human behavior at a deeper level

32:46 Levi McCollum: so they can develop the, the other technologies, whether it be like, you know, instead of a turnstile, you’re using

32:54 Levi McCollum: electronic doors, right. That are, you know, vertically closing.

32:59 Levi McCollum: That’s, that’s really what they say that they, they want to better understand so they can apply those newer technologies

33:07 Levi McCollum: to facilitate a better system. Right.

33:11 Levi McCollum: To make the system, you know, work a little bit better for the patrons who use it.

33:16 Christian Londono: Yeah, I, I, you know, a lot of the applications that we spoke about, they’re, you know, currently taking place

33:22 Christian Londono: in transit and some that are just in the infant stage, they’re of all that.

33:29 Christian Londono: And, and that’s what artificial intelligence is about is, is also understanding their behavior, understanding, you know, how people are

33:37 Christian Londono: behaving. But all this is about giving those insights to those that make the decisions.

33:42 Christian Londono: Right? Like it’s gathering all these data and it’s giving you exactly what you need.

33:46 Christian Londono: Like, okay, this is what is happening.

33:48 Christian Londono: This is, you know, what these people are doing in some cases are giving you the information right away.

33:53 Christian Londono: Like even, you know, we keep, we spoke about the buses and detecting vehicles in the bus only lane.

34:00 Christian Londono: Also heard about where it does pedestrian detection.

34:05 Christian Londono: So as a bus is about to make a turn and you know, buses, particularly buses, you know, transit buses,

34:12 Christian Londono: they have a lot of blind spots.

34:13 Christian Londono: And yes, you have professional bus operators and they’re, they’re excellent.

34:17 Christian Londono: They look all these mirrors, they’re able to do a lot of things, but sometimes every now and then they’re

34:22 Christian Londono: blocked. And, and, and that’s what a lot of times translates into, you know, running over a pedestrian.

34:28 Christian Londono: So that technology also is helping, you know, alert the driver and avoid some of the, these circumstances.

34:35 Christian Londono: So today this technology is helping agencies and I think that’s the most powerful thing to think about is when

34:43 Christian Londono: you’re looking at technology and particularly artificial intelligence as a transit professional is, you know, how can we, or where

34:52 Christian Londono: should we look to start implementing some of these technologies in your agency to benefit from those advances? This as

35:01 Christian Londono: well? Yeah, absolutely.

35:03 Levi McCollum: Well, this has been a really great conversation.

35:06 Levi McCollum: I have enjoyed it thoroughly because it gives me an opportunity to geek out on a couple different things that

35:11 Levi McCollum: these topics that I just love and I get to talk about it with, you know, one of my favorite

35:16 Levi McCollum: humans. So appreciate the time. Christian, this has been a good stop requested episode.

35:23 Levi McCollum: You know, if you please be sure to catch us next week for our all of our listeners who are

35:27 Levi McCollum: going to be back every Monday.

35:29 Levi McCollum: And we plan on doing this into the future for as long as we can.

35:33 Levi McCollum: Christian, do you have any parting words for our audience?

35:35 Christian Londono: I just want to say thank you for listening those that have been listening to our episodes.

35:39 Christian Londono: We really appreciate, you know, you following up with our episodes and giving us your time.

35:45 Christian Londono: And those that are listening for the first time, thank you for joining our podcast and we’ll see you next time.