How to do better Route Planning and Optimization (Florian Merget, Co-Founder, Greenplan)

00:00:04: Hello and welcome to the logistics tribe I'm your host boys felgendreher and in today's episode we will take a deep dive into the world of route planning and Route optimization.

00:00:15: My guest floor and merget is co-founder of greenplan a startup that was recently spun out of DHL.

00:00:21: Green pants algorithm promises to do a better job at planning and optimizing routes for Last Mile Road Freight and even field service management.

00:00:29: My conversation with Florian extended into a wide-ranging discussion around the evolution of route planning algorithms and systems

00:00:35: the promises and limitations of AI and the changing role of dispatchers.

00:00:39: Before we get started a quick nod to our supporters gray orange gray orange is a premium provider of a warehouse execution system called gray matter.

00:00:48: That uses AI to orchestrate warehouse and fulfillment operations.

00:00:52: Canal which has its own Fleet of warehouse robots but more importantly it enables you to integrate any existing robots or devices used across your fulfillment operations

00:01:01: IKEA on your desk I just two of many companies who have great orange systems in place if you want to find out more please visit gray orange.com

00:01:09: I would also leave a link in the show notes and now here comes Flora merget from greenplan enjoy hello Florian welcome to the logistics tribe thanks for being on the program hi

00:01:19: thank you for inviting me to the podcast yesterday

00:01:23: Florian give us a quick intro into a greenplan what's the short elevator pitch of what greenplan is doing yes perfect thank you so greenplan is actually a software as a service for sophisticated and sustainable to a planning we do have two or three USPS and place and number one is we can plan fully Dynamic number two is that we are considering

00:01:45: time-dependent travel times so we can perfectly

00:01:49: so the traffic for the next day and number three we can actually plan with optimal starting times and this ultimately gives our customers number one cost efficiencies and number 2 and CO2 emission.

00:02:05: Fantastic thanks for this short pitch and we'll dive into those and unpack those a little bit later in the conversation I hope

00:02:10: let's go take us back maybe to the the Argan storing the founding story of cream pie which is quite interesting because it was part of DHL take us back to to all of that part of your history

00:02:19: yeah no I'm perfect so

00:02:22: I would say like five ten years back actually Clements back man my partner here in the managing director actually started with the question of how.

00:02:33: DHL or Dodge past can really improve e-commerce right can cope with the increasing volumes and everything so at that time he was set of innovation of.

00:02:44: She passed and worked on several initiatives and one of these initiatives was actually too.

00:02:50: Set up a more sophisticated and better to a planning right so five years back then greenplan started as a project together with the University of Bonn

00:03:00: with the institute for discrete mathematics

00:03:04: why the Institute or why the University of Bonn because they are next to the MIT and several others like one of the leading institutes for these kind of problems in the world and next to this they're actually sitting right next to the DHL campus so this was something a bit and

00:03:23: yeah and then actually claim and started to develop together with the University of Bonn this Innovative algorithm with this brand new approach

00:03:33: planning in space and time I just touched it in the brief summary before we come to it probably later and

00:03:42: yeah this was actually the starting point and at that time I was not working for DHL I did my MBA in Madrid and before I was working in Consulting and then

00:03:52: had enough of Consulting and enjoin THL as Chief of Staff of the CCO and greenplan was then shifted to the CCO unit and I thought this was

00:04:03: very interesting project so I got in touch with claimants and the two of us actually thought about how we can commercialize and grow greenplan right make kind of like an.

00:04:15: Industry ready product out of it and and why because we have seen that in all the comparisons and benchmarks we did also no official tenders that greenplan delivered Superior

00:04:29: better results actually compared to all the competitors and this time was kind of like the starting point why we found it or own entity the gmbh

00:04:38: not in 2020.

00:04:40: At that time full ownership of DHL right and 100% DHL ownership but we started to build up the team in generated office customers so that's kind of like the story.

00:04:51: Yeah and I cannot I cannot the first soda pop into my or the first question that popped into my head was okay when.

00:04:56: If the intel was looking at optimizing their routes and looking at algorithms what were they using before and how different was that.

00:05:03: Because Commerce has been around and delivery has been around for a long time algorithms like that have been around for a long time so what were they what were they using before that didn't serve the purpose going forward

00:05:13: so DHL is not

00:05:16: one company right so they have several business units within several different needs you have the e-commerce you have the freight part you have supply chain with warehousing and stuff and then you have Express

00:05:28: and all these

00:05:30: different entities were using different systems sometimes manual sometimes Grand Legacy systems which actually serve the needs but we then figure.

00:05:41: Greenplan actually can do better right and really get into touch with one of the DHL units where we are currently planning over 1200 tours per day

00:05:53: with significant cost efficiency Improvement and.

00:05:59: This will standee the starting point of DHL using and leveraging greenplan

00:06:04: yeah yeah yeah so you're officially launch 2020 you said I know it's 2022 so for the last two years you've been inside of DHL and where you only serving

00:06:14: DHL internal customer so to speak or we also expanding and acquiring customers from the the outside world so to speak Uncle question

00:06:23: obviously first of all I'm just DHL internal and then we actually get the

00:06:28: corporate board decision that we also can go external and started to approach also external and express providers facility management providers Pharmacy delivery companies so all kind of companies who have kind of like

00:06:44: shipments on need to plan tours and they're actually.

00:06:48: Yeah we really started in 2021 I would say to really go out and to the market and reach out to

00:06:56: point of time still external customers

00:06:59: yeah and now you're a completely independent so you're 100% yes owned by yourself and your partner I guess if that's understood that's correct yeah so exciting times

00:07:09: it's super exciting I can tell you it from from a consultant from the consultant to the real world of actually actually doing business not just talking about it to to an entrepreneur yeah how's it been so far it's.

00:07:21: It's very exciting it's very intense on the other hand right because you're kind of like dealing with

00:07:28: every every topics from marketing HR sales customer implementation idea research development investor search so everything's on the table but.

00:07:39: Super exciting it's a very exciting Journey we are can't lie

00:07:44: building up the team very quickly because in large corporations it's not that easy to get resources so we now have more freedom to really build up

00:07:55: the team to grow and greenplan more aggressively.

00:08:01: And on the other hand we have seen that with this

00:08:06: management buyout of greenplan out of the HL that here further external customers or leads are.

00:08:13: Also approaching us now so before they were kind of like hesitant okay DHL is kind of the owner but now with this 100% independency of the edge as external customers really come to greenplan and and approach us and

00:08:29: yeah and that's quite a good story and how big are you right now

00:08:32: what are you starting starting from with so we are 14 people in the core team and then we have 10 people on the development sites on the University and another like 8 to 10 people on the it

00:08:46: I'm development site

00:08:47: yeah and you tease a little bit earlier sort of you started describing so the the evolution in the history of algorithms and yeah and how you innovate and develop something completely new that that wasn't there before to your claim so take us back to

00:09:01: sort of the history again like a comprehensive history of.

00:09:05: I love how these types of systems came about and how they evolved over time and where we are today 2022.

00:09:12: Yeah no they're very interesting question and they're actually happened a lot during the last 5 to 10 years so I would say like.

00:09:22: Everything started kind of like with manual to a planning right so I'm mainly excel-based or kind of like table based and and then it went to more semi automated tools where the dispatchers or the planners who actually are responsible for planning the to us got some kind of

00:09:39: decision support however during that times the power and heart of the tool planning still.

00:09:48: What's the responsibility of the human being of the dispatchers themselves right

00:09:53: and and kind of like they are knowledge they knew for example OK is a track able to cross the bridge is a track able to go this direction or that direction and all these kind of knowledge was

00:10:06: implemented into this semi automated planning tools and.

00:10:12: Then actually more or less like the algorithms came around the corner like mathematically optimized to planning fully automated or up to fully automated to a planning and.

00:10:26: Yeah we're actually systems kind of like outperforming this this manual planning right the algorithms.

00:10:36: But still these algorithms are dependent on the dispatches knowledge for all the different business rules for all the processes that have to be incorporated and side note this we are including also in Greenland so our.

00:10:51: Greenplan algorithm is super smart and super sophisticated.

00:10:55: Planning all the tools however the customer specific rules like if for example we are talking about an Express customer with different time Windows priority deliveries this has to be included into our system and.

00:11:10: Knowledge provider is actually the dispatcher or the planner in the end and next to this next with this algorithm based.

00:11:19: Tour planning towards there's also kind of a stream of AI based applications which are

00:11:26: popping up here and there and everywhere and we have seen and this also proven is very interesting that AI is actually not necessarily applicable for this

00:11:39: complex field of to planning because AI is usually looking for patterns right and then optimizing these patterns.

00:11:49: In this kind of delivery process you have like always

00:11:54: different deliveries each day so Monday is not similar to a Tuesday right because n consumers actually ordering stuff differently and this is why the two was actually looking

00:12:08: different each day and AI is

00:12:11: at the moment not capable of kind of like outperforming the algorithms.

00:12:18: Just to give you an example yeah let me let me just quickly an ejector I mean isn't

00:12:24: but isn't the claim of people advocating AI systems that eventually all the knowledge that you for example any dispatcher has about a particular route or particular region can eventually be learned.

00:12:36: Through looking at patterns right so just like a human being a dispatcher has learned patterns over time in a I can do that but

00:12:43: a lot more efficiently a lot more at Large Scale than any human being could ever could ever do like what's how do you address

00:12:50: those correct that's very good I mean you mentioned the perfect word which is patterns right and we have seen that there is not similar patterns.

00:13:01: From a Monday to Tuesday to Wednesday so this is why I actually I have a rhythm based approach here is more efficient and we have also seen this in the most recently Amazon Last Mile Challenge.

00:13:14: Each so a

00:13:16: professor of our R&D team took actually part in this Challenge and together with his team and the greenplan idea actually won this challenge by 42% and to the second place and the second place was the MIT

00:13:34: and they actually kind of like try to solve this problem AI based right and so here we have a

00:13:42: clear difference of what algorithms can do compared to a i based solution at the moment.

00:13:49: Okay maybe that's it that's a good point to to kind of separate the two just so that the audience understands the clear difference between what we mean when we say in AI system versus algorithm because

00:13:58: oftentimes in typical language these two things get completely messed up my mashed and like sort of confused and used in the same way so what's the.

00:14:08: How do you delineate how do you

00:14:10: differentiate between what's an algorithm and what's actually an AI or is the demarcation the line not as clear as we think yeah well I would say phone algorithm kind of like we

00:14:20: developed really Akia algorithm or a function of how to optimize the.

00:14:28: Data that comes in every day according to several constraints I would say

00:14:35: fight and so.

00:14:38: Our algorithm is kind of like Optimal optimizing in the best way the data that comes in and as an output gives kind of like the best tools

00:14:48: possible AI patterns compared to this kind of like try to learn like.

00:14:56: Yeah so if you have 100 shipments on a Monday which I delivered 200 addresses and then you have on a Tuesday 150 shipments address

00:15:07: delivered 250 addresses then the AI based brain actually

00:15:13: tries to build some patterns out of these out of these delivery days right and tries to kind of like

00:15:24: put it as a blueprint on the next day on the Wednesday on the Thursday.

00:15:29: This due to all the changing addresses and customer requirements and everything is not

00:15:37: possible are not necessarily possible at the moment so you can't get that much efficiency gain out of this

00:15:46: pets on based blueprint logic compared to a daily new algorithm based calculation of optimal tools

00:15:54: yeah and does this apply to all kinds of deliveries or does it apply mostly two capsules of parcel delivery Last Mile

00:16:02: because where there's lots of other tools and lots of other route optimization and full truckload less than full truckload large distance so talk to us about how a i versus algorithms work better in some use cases where there's others

00:16:14: I think where you have like very same patterns each day like we have for example I don't Know full truckload B2B I'm delivery so you're going everyday to the same location

00:16:28: probably AI is very.

00:16:30: We as greenplan that's actually very good question from Uruguay is greenplan are really focusing more on then the LTL or PTL and less than truckload Patrick.

00:16:39: Which go actually around and don't have like similar patterns every day and it's applicability is much better than for women.

00:16:48: Definitely yeah

00:16:49: and but since we're we talked too much about DHL earlier so this is not only applicable to Last Mile passive delivery but it's actually more right yes yes.

00:16:58: I'm so we are currently focusing on several Industries number one is the parcel express and e-commerce obviously so mainly Last Mile yeah it's one of the most demanding Industries I don't have to tell you

00:17:12: how do you know how customer requirements are evolving with like at all

00:17:19: getting in same day and all these kind of stuff so here actually we need some very good to a planning Solutions

00:17:26: makes this actually what I just touched before the roll trades mainly LTL or PTL and here we actually supporting our customers in there

00:17:37: just in time transport especially at specifically when it comes to these complex LTL Transportation networks so here greenplan is very valuable another one is retail.

00:17:49: Where we actually supporting retail companies for home deliveries and a brand new area is filled Services it's very interesting because we are also planning

00:18:02: technicians who are out there in the field I need to maintain for example assets.

00:18:06: All right so it's kind of like similar the problem technician goes out as like I don't know five ten jobs a day I'm you have to.

00:18:15: Optimize the two away he's driving you have to consider all the different handling times the skills of the technician so this is also very very much.

00:18:26: Coverable and within greenplan.

00:18:28: Yeah let's let's let's focus on for a second on the on the freight forwarder so speed it's your own in German that's right truck companies crunch our companies delivering.

00:18:37: So when you go into these situations and you're trying to acquire a new customer what do they commonly work with so they have two spatulas right and everything is done manually still or other Ewing with transportation management systems what's the current set-up that people use when you come in there it depends it really depends but these are the he actually did it

00:18:52: the two passwords here number one and

00:18:54: either they do like everything fully manual or extra based number two they have TMS transport Management Systems in place and.

00:19:04: So we have one customer and place actually and.

00:19:12: We compare then our system in an official tender and compared to there

00:19:18: previously system and were significantly better in terms of

00:19:24: cost efficiencies so five companies two parts and the second-best was actually eight percent behind us and this then was the kind of like starting point for greenplan to be implemented

00:19:36: within this customer however here I'm coming back to the TMS because TMS is not.

00:19:44: Specifically about 20 planning TMS is much more right you have the truck and trailer track and Trace and you know everything is kind of like.

00:19:53: Write a lot of functionality but but but to our planning is a corporate right so every every decent

00:20:00: TMS does have some functionality for for about 20 minutes so that's what you're competing with we coming in there somebody has a TMS or somebody has like a really experienced and you know like a dispatcher who believes that

00:20:11: the knowledge that they acquired over time is that there's the gold standard so to speak so it's not in Oregon easy sell right correct no it's not an easy said no and this is why we usually actually saying.

00:20:22: Kind of like we're approaching customers with the so-called Benchmark calculation so that the customer can give us like for I don't know a day or week of

00:20:31: out of one Distribution Center have four Depot I'm data we then put it in our system yet like the optimal greenplan to us and then the customer can compare it with the

00:20:40: existing to us and then actually can see the differences and the advantages age.

00:20:45: Yeah it is the idea that your system will sort of integrate with the TMS system would it replace certain functionality to talk to me about how your system works with other existing systems out there.

00:20:58: Again two streams here number one if the customer has the TMS and then we are more than happy to be integrated as the routing engine as the heart into the TMS

00:21:08: we have a rest apis in place so we can connect very easily to transport management system on the other hand it's also very good for the dispatchers or planners because

00:21:19: they still have the same user interface

00:21:23: Sprite so it's just in the background they're getting them from greenplan better tourists but still have the same user interface so.

00:21:32: Change management is not that

00:21:35: heavy here in this case yeah but you still have to I would have imagined I'm just going back to its great algorithm but garbage in garbage out if you know the best algorithm fed with

00:21:44: terrible data is going to lead to terrible results so it's it's super important to get the data right to get Integrations right so the data actually goes into it

00:21:54: it's trustworthy as reliable as clean is accurate all of that talk to me about how how you make sure that that actually is up and running how do you assist.

00:22:01: I mean if you have a like a small midsize Freight forwarder that's been doing manual work for years or decades how you get them to really

00:22:09: you know that's a desert digital transformation Journey even if it's a small one right but but talk to me all about how that works and what the expectations there are four

00:22:17: for these types of customers yeah now that's I mean we are also facing these kind of questions right with customers and

00:22:25: Yeah you mentioned it kind of like digitalization and going more into digital products is actually key here and number one is definitely the data right so first of all I mean we we have a team then

00:22:39: at the customer site and going number one into the processes and number two going into the data actually discussing

00:22:47: the customer data format data inputs everything that we would require clean it and then based on this actually customize our model to what's the.

00:22:57: Usually what we are to be more concrete ER precise what we're requiring is it

00:23:03: two types of data number one is the master data which is all the vehicles different vehicle sizes capacities volumes and speed profiles the Depots and the drivers.

00:23:17: Let's say different working time models

00:23:19: this is the one the question there is this typically the this master data is it typically data that can easily be exported from some sort of TMS or maybe Erp system or is it something that needs to be manually collected somehow but it's also an easy transfer usually it's there

00:23:31: usually the data okay right I'm and then then the second data type is the so-called transactional data which comes in

00:23:39: every night or every day and this is then actually the ship

00:23:42: that's right the shipment data with all the time Windows priorities handling times weight and so on and so forth and here actually the

00:23:54: data quality has to be right in order that greenplan can calculate the optimal results just.

00:24:02: To give you one challenge here or to mention One Challenge here which we actually faced is

00:24:08: the geocodes right so geocodes have to be precise as precise as possible so based on Geo codes actually greenplan is calculating the tours and if the Geo codes are not very

00:24:22: accurate then obviously what you said garbage in garbage out right.

00:24:27: Yeah so this is this is an this is individually different from customer to customers have some some got their geocodes right some not so right so that's a gradient it's a yes it's scale so to speak yeah okay right.

00:24:38: That's there are more and more systems out there who are actually focusing on this problem right.

00:24:44: Yeah I'm just a man named a few on the one hand you have Google Maps right where you can actually I'm kind of like do all your geocoding then they're what three words for example and so many others who are actually I'm kind of like focusing on this problem.

00:24:58: Yeah.

00:24:59: Yeah can we Deep dive little bit more into the algorithm what actually goes into it what data points does a collect and how does it calculate routes based on it and then what's the what's the actual output.

00:25:09: It was a little more insight into the actual algorithm.

00:25:12: Know what I mentioned before like number one is the master data that has to be has to be.

00:25:20: Like they are for the algorithm to be able to work which is the fleet

00:25:26: right with the vehicle size and everything and we need the Depots and the drivers with their working times so this is kind of like the fixed data right which is are everyday and then you have the transactional data which comes in on a daily basis with all the shipments and if greenplan has these kind of data then

00:25:45: then greenplan.

00:25:47: Cake relates the optimal routes how does it calculate the optimal routes so data comes in and then usually our algorithm for a let's say.

00:25:59: 50 tour Depot or Hub takes like 30 to 40 minutes to really calculate the optimal tools and then actually as

00:26:09: and output gives out.

00:26:11: Towards the stop sequence with the different handling times with the driving times with the considered priority deliveries Etc so all these kind of data actually goes then into

00:26:24: customer systems

00:26:26: yeah and then as the route unfolds the and things happen as always traffic Dan shows up you know whether weather patterns Etc it automatically recalculates the routes with this

00:26:38: it's real-time recalculation of the routes as things happen

00:26:42: does it take real-time information into consideration and good question I'm so we started let me just

00:26:48: give me a few minutes to kind of like explain where we are coming from so and then I come to the real-time thing so

00:26:56: our first cases actually came from the postal area right and usually kind of like.

00:27:03: Shipments come into a half in the morning and then to escape calculated then the Sorting takes place and then the tour's go out right the vehicles go out and deliver all the all the passwords which means that the

00:27:15: twerk a collation has to be done like at you know 3 4 5 a.m. in the morning

00:27:21: 4-H lead the next day right and.

00:27:26: This is actually our Focus so we are calculating the tools during the night and then giving as a nurse

00:27:34: results the tours to The Postal operators freight companies Etc who will then drive the tools during the day.

00:27:43: Right and here we consider historical three profiles from TomTom in this 5 minute intervals we know how the traffic will look like on a Monday on a Tuesday and Wednesday Thursday Friday.

00:27:57: A certain point of time at a certain street so you have an average.

00:28:03: Velocity of flow velocity there and this is considered in our system during the day then when the drivers actually out on the road and driving

00:28:13: usually we have seen that they are typically using then Google Maps or other systems which considers then

00:28:20: accidents or a talk traffic jam or whatever and then kind of like it just this however

00:28:29: we are currently developing also a feature where we actually can include this ad hoc events in the greenplan.

00:28:37: To a plan as well so that greenplan will then recalculate the optimal sequence after for example and they talk at event or whatever might have happened during the during the course of the tour.

00:28:50: Yeah so well that means that

00:28:53: in the meantime for now there is no dynamic recalculation or change of the route as unfolds okay so now it's once it's calculated once it's delivered to the right

00:29:03: for it to the driver he goes out there and if things happen they happen and he continues to he or she for continues on them on the route as it is yeah okay.

00:29:13: Yeah what do you what do you say to people when they say well things always

00:29:17: change I mean stuff always happens every single day there's a traffic jam and have to switch routes and have to change plans and I'm stuck with this plan that assumes that everything is going is going according to plan how do you how do you deal with those criticisms.

00:29:30: Yeah I remember these are some of the questions that we are

00:29:33: actually getting and so because we are having this very precise predictive data of TomTom usually like the traffic jams and everything is already covered

00:29:45: right within our to a plan so for example then greenplan nose

00:29:51: yeah Ron Colonia have a lot of traffic jam in the morning during rush hour okay we are avoiding this this highway or that Highway because this is too much traffic there right so greenplan knows this already during the day

00:30:05: um if there is an accident for example I'm or whatever the drivers usually know what to do right they kind of like

00:30:14: take then Google Maps and where's the fastest route or take another navigation system and then actually go to the next stop so this is not very.

00:30:23: Very very critical from what

00:30:25: from what we have seen okay but there are other systems out there that completely I mean for example other possible providers they have their own internal systems that they've developed over time and

00:30:35: it's actually exactly that we're it's a it's a calculated route when the route starts but then it gets dynamically updated as things unfold and you don't leave stuff up to the individual driver to make decisions but here

00:30:47: I understand there's still an element of human and decision-making I'm part of the driver correct great to to actually adjust the route as necessary.

00:30:57: And but then the next step you you alluded to it there will be a next evolution of it where you also take.

00:31:03: Evolving situations into account how work that how would that unfold so

00:31:07: yeah so that we are more really like focusing and can react to real-time events right

00:31:14: I'm and just to name a few so sometimes.

00:31:19: When you're out on the road and have a lot of deliveries and sometimes pick up events come in

00:31:24: right so these pickup events all returns have to kind of like.

00:31:29: And considered within the route planning and this is what we already from an algorithm perspective developed but I can't lie rolling out in our standardized industrialized

00:31:41: I'm so so this is one one thing that we are doing here for example.

00:31:46: Yeah and we've talked a lot about the algorithm but what is it what is it actually it's a piece of software and it runs in the cloud and its software as a service and how does it show up in my my systems I mean is it a standalone.

00:31:58: Software application that I open up to do my planning or is it something that shows up in sight my TMs or how do how does it work in actual practice Yeah

00:32:07: it depends on the customer needs so number one we can integrate into the TMS right I'm out there I'm so we are just the routing engine and then the user can always comfortable with

00:32:19: with the TMS user interface or everything's fine on the other hand we also offer greenplan including the user interface right so it's

00:32:27: based in greenplan shows all the tours on the Gantt chart and everything so very user-friendly and very nice so this is what.

00:32:36: Can also provide an offer to our customers.

00:32:38: Yeah and then does it get played into the individual systems that the driver see I mean there's some sort of mobile element to it and stuff the the interface type drivers use inside the vehicles yeah

00:32:49: how does that work so usually drive us have their handheld devices right sometimes with navigation or whatever and as greenplan actually as one of the major

00:32:59: puts of greenplan is the stop order with the different timings and so on this can then be.

00:33:05: Imported or transferred to the drivers and head so that he actually can see okay stop a is next or B follows then subsea

00:33:14: is the is the further one so there's two elements is element that the driver sees and there's an element

00:33:19: dashboards for example the dislikes betcha yeah in the end for the dispatcher and as you keep using the system day after day after day accumulates data

00:33:28: yes what possibilities do I have to analyze how good I've done my routing for example what sort of elements of analysis you have mmm

00:33:36: so you mean kind of like this Statistics and kpi is right and they come out so I've been using the system for four year and I want to pull up charts where it shows me how well

00:33:46: my predictions have been and how accurate

00:33:48: and where I've saved km costs you to your name it and these I say actually actually the things that customers are looking for a number one they want to see right reduce km so really compare it to like

00:34:05: what they previously had or also a is and to be comparison right so usually if drivers go out and then they have Scan events right and so you can really.

00:34:17: Compare the scan events with what greenplan calculates with what was there previously so you have like

00:34:23: I'm three steps here this can be compared and number two is what's getting more and more important CO2 emissions and sustainability.

00:34:32: Because of reduced km reduced vehicles used ultimately you are reducing the CO2 emissions as well and this is what.

00:34:39: What we can show here as well on our standardized dashboards.

00:34:44: Yeah and what additional data do you need to accomplish that I mean if you only know what what vehicles and where the distribution centers are where the drivers are what the capacity constraints on so forth but you need additional element of

00:34:55: CO2 emissions per vehicle permit correct what have you yeah how does that come into play where does that where does that gets was from.

00:35:02: So usually they're kind of like standard values right per vehicle type and whatever and then you can.

00:35:08: Not very easily but very straightforward like do the calculation and do the math in the end.

00:35:13: Yeah well we are currently discussing is also electric vehicle.

00:35:19: Right which we are also considering in our planning here there is a bit more of a element right because it's not just the reach but you have also

00:35:27: to consider the weather influence the topography right if there's a lot of hills and then obviously the electrical

00:35:36: vehicle reach will go down and and these things we are currently also kind of like developing and considering

00:35:43: and what's the what's the business model I mean do I pay by user by km what's the what's the model so it's software-as-a-service so I can yeah I could try it out it's you know it's a little risk it's low barrier of Entry I hope

00:35:54: once you've done all your data right talk to me about the business model how do I how do I pay for this

00:35:59: so we are a licensed based or we have a license based business model and it's fully value-based so you.

00:36:08: Half or you're paying per calculated tour and the.

00:36:13: So if you have like I don't know 100 tours a day then you pay a certain Euro amount for each of

00:36:19: of these hundred tours and the Euro amount depends actually on the volume of the customers right if you have a very small customer than obviously it's a bit higher it's around for 25 Euros per calculated tour if you have

00:36:33: very high volume and a lot of tourists then it can go down to one Euro tour.

00:36:40: Okay okay well I appreciate you naming it actual numbers yeah of course I mean it's good yeah transparent yeah right why not yeah

00:36:47: yeah and how do you if someone's in the audience interested in trying it out what's your what's your typical method of onboarding people giving you the try what's the what's the preferred method to

00:36:56: to see if it's a good fit there yeah

00:36:58: so usually we are doing this kind of trial or Benchmark calculation where a customer would send over a data of a typical week of a typical day poor Hub and then we actually feed it into our system see if all the data's right format is right and if we considered all the requirements correctly

00:37:17: and greenplan will then calculate we actually present the results and to the customer and if

00:37:24: he or she is then convinced then obviously we would go into kind of like a first project phase Where We Are.

00:37:33: Together with the customer capture all the business Woods processes data put everything right in place and then go into a proof of concept

00:37:43: usually a few tours a few days in one of the depots.

00:37:47: To see if everything is right it just tweak our system make it it ready with all the connections and then roll it out

00:37:54: and that's kind of like the straight forward some very easy process here yeah and you mentioned earlier that your customer base or the types of customers and industries are quite

00:38:03: there's a wide range yes but what are the typical sort of the core customers that you really going after that really should be taking a look at your your system.

00:38:11: Definitely parcel express any Commerce because we can

00:38:16: very good consider all their requirements and have a very powerful algorithm then Freight and mainly LTL or PT l-- because there's

00:38:24: much more optimization potential here than full truckload and what I mentioned before I'm Field Services.

00:38:32: That companies out there that have a lot of let's say technicians or field sales or whatever so this is what we can plan perfectly

00:38:40: just to give you one custom example you may be so we of course yeah sure so we are planning for a German facility

00:38:49: Tea Company that funded technicians out in the field looking after assets right to maintain these assets and before greenplan they were able to do three jobs per day per technician

00:39:00: with greenplan it's going up to six or sometimes even seven jobs per day per technician.

00:39:08: Based on better to a planning and based on consideration not for all the flow velocities and everything and this is significant increase

00:39:16: yeah it is what's the what's the comparison there before they didn't do any type of planning so this is coming from from nothing to to green plant has to come from an existing system to greenplan within existing system very good dispatchers and place already I'm who actually planning this but

00:39:31: thanks to really like the power of this algorithm and I said it I think at the beginning it's more than 50 years of mathematical

00:39:40: expertise and knowledge which went really into the development into the research and development of this algorithm and that makes the algorithm so strong.

00:39:51: Yeah and is the ambition to replace human beings so dispatchers I mean that's probably one of the questions always shows up because there's going to be Dispatchers in the room when there's a decision being made of what systems used in the future is at been an issue I mean are you

00:40:03: replacing people or is your pitch.

00:40:06: Well we're not really replacing them we're free them up to do something more higher level you know that's that argument that that vendors of that type of system typically use no I mean it's not about

00:40:18: bus

00:40:19: wedding here so I'm obviously we are taking this patches into consideration because greenplan the greenplan system has to learn and to be set up the proper way so that it operates the proper way

00:40:33: right and for this we need this kind of let's say

00:40:37: digital FrontRunner of dispatchers who would be willing to really support and set greenplan up in the right way

00:40:45: so that we're considering all the customers specific needs and rules and everything so this is why we need this patches on the other hand of course if you're talking about cost reduction fully automated systems

00:40:57: ultimately would reduce the you know the people are required for planning towards.

00:41:05: Yeah and of course any smart dispatch oh well we'll see right through it and think okay I'm setting the system up and the next thing you know you know I'm

00:41:12: my job is redundant because the system is doing it better than I have so do you see some level of skepticism amongst the crowd of dispatchers out there who are

00:41:20: getting potential displaced by the stuff yeah well usually there is some kind of change management and

00:41:27: communication is always the key right of how to really kind of like Get the.

00:41:35: Dispatchers involved in such a way that they are supporting the implementation of the greenplan system and this is this is very very important here right and so if the dispatchers actually say no we do not want the system.

00:41:50: Obviously we don't have the chance right to replace or to kind of like improve their job so we really have to make sure that we get

00:41:57: dispatchers on board and on the one hand really offer them that they can set up the system

00:42:03: with us to get on that they can maintain it and on the other hand Freedom up for more strategic initiatives so scenario planning or whatever

00:42:11: so these are kind of the strings here.

00:42:14: We are following yeah and I am mostly focusing on Germany right now are you also going europe-wide or even internationally what's your what's your international plan look like global.

00:42:23: All right come on ambitions

00:42:25: you like it it's about the mean of course you know your small team you know you can't scale from from 0 to 100 in notes I mean you gotta prioritize in some way so how would that look like we already have some streamers in Germany at the moment and in Europe

00:42:40: in northern Europe in southern Europe and now reaching out to the UK we do have currently

00:42:47: Pilots or first calculations running in Asia.

00:42:51: And in the u.s. so I'm in the long run definitely we want to also set up our offices in us and in Asia to really have like yeah kind of like Market presence here as well

00:43:05: in these regions.

00:43:06: Yeah you are you actively looking for investors them because that's going to that's going to be expensive yeah right International expansion is a is a tough job for any small start-up and needs to be financed

00:43:15: yeah you're looking at financing right now are you do you know you're independent you're outside of THL details no longer an owner so

00:43:22: no I were what's you investing yeah I would say money is always good to grow right so that's why I definitely in the.

00:43:32: It's mid short to mid-term we are looking for further investor funding and so that we can really scale and tackle the markets and make our customers happy

00:43:42: worldwide yeah I mean so far last couple years the the climate for investment has been good I mean a lot of projects have been funded and funded with amazing amounts of money so who knows how long that window is still open so good luck good luck on your thank you and your financing Church.

00:43:57: Florian thanks very much for joining us today was very good inside deep dive into that type of situation and problem and then cello Matador we haven't talked about it yet and I appreciate you tackling all the tough questions

00:44:10: good stuff thanks a lot of your time thank you

00:44:14: all right that was the logistics tried podcast episode with floor and merget from greenplan if you enjoyed Today's Show please subscribe to the podcast so you don't miss any of the future episodes I'm bored felgendreher until next time.