2020 STEX25 Accelerator Startups Day 1 - Startup Lightning Talks with Q&A, Session 2

MARCUS DAHLLOF: Our next speaker is Daniel Theobald, founder and CEO of Vecna Robotics.

DANIEL THEOBALD: Thank you, [INAUDIBLE] It's great to be here. My name is Daniel Theobald. I'm founder and CEO of Vecna robotics and co-founder and president of MassRobotics. We are focusing on autonomous material handling, primarily bulk handling. So think about pallets, large items, et cetera. Next slide.

As you look around the room you're in, what you might realize is that most of the items you see spent a significant portion of their time pre-store shelf or pre-online purchase on the pallet. And the supply chain really is as we've seen recently, very sensitive to disturbances. COVID has made us more aware of that than ever. And there are a lot of problems that you run into-- safety issues. It's hard to hold on to those supply chain employees. A lot of inefficiency in the supply chain that leads to non-ideal throughput, a lot of waste.

Many people don't realize in the warehouse-- some research recently showed that up to from 30% to 50% of a worker equipment's time is either unutilized or underutilized. And there's a real problem dealing with problems. They call these exceptions. What do you do when something goes wrong? So the need to move things efficiently from loading dock A to loading dock B can encounter a lot of unexpected challenges in the real world. Next slide.

So what Vecna robotics has done is built Pivotal, the world's first day AI SaaS based orchestration engine. And part of that pivotal platform is what we call the autonomy kit. The autonomy kit is essentially a black box that can turn any piece of industrial equipment into a fully autonomous mobile robot. So think driver-less forklift, anywhere you're moving pallets or large items in a factory or a warehouse or retail operation, that can be done completely autonomously, autonomously and safely.

But the fact of the matter is that once you've made that piece of equipment autonomous, the real value that we can unlock comes from orchestration, making sure that you've got the right piece of equipment and the right human worker in the right place at the right time. Next slide.

And it's really all about learning and adapting in real time. None of these operations are static. It used to be that you could sort of plan out a warehouse and have it operate largely unchanged for a decade. But these days, it's hard to predict what you're going to be doing even three months from now. And what that means is that you need more flexible technology. You need systems that are not built in heavy infrastructure. And the autonomous mobile robots can be a really important part of this.

Here you're seeing a autonomous pallet jack moving items around in a large warehouse, saving a significant amount of time. But some really exciting things happen when you bring robots and humans together in a very coordinated way. And there's a use case here. FedEx, where you can see that when we had just the robots installed, they provided great ROI. But then when we started coordinating the robots routes in real time based on the ground information pulled in from the system, you got a large increase in throughput. But then when we coordinated both the humans and the robots, we got over two times increase in throughput. And that's really again this idea of having the right resource in the right place at the right time. Next slide.

MARCUS DAHLLOF: Daniel, we have to almost wrap up soon.

DANIEL THEOBALD: So what are we looking for? Partnerships in a number of industries. If you have a piece of equipment that you need to have automated as an OEM, feel free to reach out to us. And obviously if you are a warehouse or manufacturing facility moving pallets and other items around we'd love to talk to you.

SPEAKER: We'll take a couple of audience questions. So there's other companies like [INAUDIBLE] that have some similar orchestration engine. How are you unique?

DANIEL THEOBALD: Yeah, we have been focused on solving this problem for a couple of decades now. And that I think one of the biggest things that separates ours is the ability to really operate in real time and pull. We track where all of the equipment is. We're able to track all of the human workers are. And then Pivotal is sort of like a Grandmaster chess player. It understands what demand-- the demand generator is asking for instance, your warehouse management system or your manufacturing execution system, your order management system. And it understands what resources are available and where they are.

And it is able to make the ideal assignment of tasks. It's somewhat like an Uber ride sharing algorithm in that you're trying to coordinate a large number of resources in real time. And it deals very robustly with errors and exceptions. So plans never go as expected. And the system's ability to do what we call iterative plan and repair. It's got a plan. It's executing the plan. The plan doesn't go as expected. It replans in real time, really sets our system apart.

SPEAKER: Great. Marcus, do we have time for one more?

MARCUS DAHLLOF: Let's take one more please.

SPEAKER: How do you track the human workers? And also how do they feel about having-- adopting robots on the floor?

DANIEL THEOBALD: Yeah, it's a great question. The tracking of human workers happens through a number of different technologies depending on the customers, anywhere from ultra wideband to Wi-Fi based. But it can also happen based on their most recent scan. The workers, I think you see oftentimes in the media the whole robots versus workers thing. So there's some apprehension before they start working with the robots.

But as soon as they start engaging with the robots, the very typical response we get is wow, this makes my job way better. And if I learn how to use this, it's going to help my career because then I can go and teach other people as well. So it moves from a threat to a career enhancement very, very quickly.

SPEAKER: Thank you.

MARCUS DAHLLOF: Thank you, Daniel. Let's go to the next speaker, Daniel Wentzloff co-founder and co-CTO of the Everactive.

DAVID WENTZLOFF: Thank you, Marcus. Actually hi everyone. My name's Dave Wentzloff. And I'm the co-founder and co-CTO at Everactive. Let me share my screen here. At Everactive we are developing full stack monitoring solutions for industrial applications, including self powered wireless sensors that are maintenance free combined with cloud based software to provide insights in real time from that sense data.

A typical industrial facility will have thousands of assets, including things like motors pumps and valves. And today these assets are to a large extent just unmonitored. And at Everactive, we've developed a continuous monitoring solution, including wireless sensors that operate from harvested energy with no battery. This allows us to deploy lots more sensors that are all maintenance free without having to worry about going back and changing a battery. And then we operate our sensors continuously.

They're always monitoring because they have a virtually unlimited power source. And that provides a rich set of data to the cloud. These new data streams allow us to derive information on specific assets such as the operational state of a steam trap or flag if maintenance is required on a motor and then deliver those insights directly to our customers in real time.

We provide an end to end solution which we sell as a service based subscription today. This means we provide everything, including the industrial grade sensors, the communications thorugh our own wireless network that we manage, which then backhauls to the cloud where we securely store the data and run our further analytics on it. Our key differentiation comes from our custom ultra-low power chip inside each of our self-powered sensors. The chip manages the energy harvesting, the sensor interfaces. It does local processing-- does all the wireless communication, including an always-on receiver.

And our wireless network can scale to thousands of devices per gateway, each with millisecond latency and a 250 meter range for our latest product, so a couple of football fields. This allows us to monitor thousands of assets without adding a single battery to the maintenance schedule. We currently offer two products, a steam track monitoring service that's been in production for about a year and a half and a machine health monitoring service which is currently in beta with some existing customers. But we'll be releasing later this month.

Both products are focused on key paint points in industrial processes where there can be thousands of traps or motors spread across the factory that are rarely inspected. When these do fail, they immediately start costing real dollar losses through wasted steam or poor efficiency. And sometimes these failures can go undetected for months or years. Our continuous monitoring provides 3 to 9 extra return on investment every year with a payback period that's typically measured in months by reducing the cost of downtime, energy, maintenance, and safety.

We do recognize that the IoT is a gigantic market and are interested in partnering in order to bring this technology to even more applications. So today this can look like a continuous monitoring solution that leverages our current flexible sensor platform and wireless network to target new industrial assets. And our current sensor platforms can support all these harvesters and sensors that you see here on the slide. And then once the data reaches the cloud, there's additional partnership opportunities because there's virtually endless possibilities for what we can do with that data.

We're also looking at the next generation of our low power chip to enable battery-less monitoring with smaller size, longer range more compute, and additional sensing capabilities. And we'd be interested in discussing how we could partner to help shape our future platform to deliver tomorrow's IoT solutions as well. So I'll leave with you a few questions to think about just to see the discussion on how we make partner.

What do you want to monitor that is currently inaccessible for some reason? Where would you put a no-maintenance battery-less sensor? And what do you really want to know about your facility or operation? Or perhaps you might want to talk about how you could take advantage of one of our existing steam trap or machine health monitoring services in your facility today. So thank you for your attention, and I'm looking forward to continuing the conversation.

SPEAKER: Thanks David. We'll do a couple of questions from the audience. How does a connectivity deployed in a factory?

DAVID WENTZLOFF: So in a factory, we will install-- with help from our customers, we will install gateways. Those gateways talk down using our ever-active wireless sensor network to our battery-less sensors and then backhauls to the cloud typically over LTE. But we can also support Wi-Fi and ethernet.

SPEAKER: Great, and how do you go to market? Can you elaborate a little bit.

DAVID WENTZLOFF: Yeah, we're taking this to market as a service. So it's an annual subscription on a per sensor or per asset basis. That includes all of the hardware. It includes the gateways. It includes the LTE backhaul. It includes all of the cloud infrastructure. And it includes developing those real time insights directly to our customers.

SPEAKER: Great. How do you scale this when you have support- specific applications like stream stress? Can the customer develop their own?

DAVID WENTZLOFF: Yeah, I love the scaling question. We thought about scaling from the start. The IoT is expected to reach a trillion devices. That's going to require scale. So first just for our system, our system is designed to support 1,000 nodes per gateway. So we can scale to thousands of nodes in a facility, each with low latency and continuous communication to the cloud.

So as far as new products or new assets that we may monitor, we are working with our existing customers to identify what those maintained ports are. I'd love to talk to anyone who wants to follow up with a conversation about what they might want to monitor.

MARCUS DAHLLOF: The next speaker, his name is John Wass, CEO of Profit Isle.

JOHN WASS: Great. All right, well, thanks everyone for the opportunity to speak today. I'm John Wass. I'm the CEO of Profit Isle. My partner is Dr. Jonathan Burns, who's been teaching at MIT for the last 30 years. Our company Profit Isle is based on a book Jonathan wrote, Islands of Profit in a Sea of Red Ink. And we have taken the concepts of that book and created software that allows companies to look underneath their profit-loss statement, down all the way to the-- [INAUDIBLE] fast, down all the way to the individual transaction level.

And by looking below the profit-loss statement, we're able to see patterns of profitability that are not available to companies in their traditional financial reporting systems. And we have consistently been able to help companies improve their profitability by 10% in less than 12 months. In today's age, the speed and complexity of business are really accelerating quite dramatically. And things like precision marketing are really driving the organizations. But many companies haven't really looked at their internal systems as well as they need to, specifically the P&L really hasn't changed dramatically in almost over 200 years.

It was designed by humans for humans and in an age when averages were sufficient because prices were uniform, and overall complexity was much lower. In the digital age, you really need to see below that. And you have to see exactly the profitability for each of every single transaction. And the only way to do that is with a decision support system like Profit Isle. We have put about $100 billion of revenue through our model. And we can say with confidence that the old paradigm of gross margin does not correlate to profitability.

And now, especially in the COVID pandemic, where manufacturing costs are changing so dramatically, standard costs are unreliable-- where distributor costs are really unknown given how dramatically their supply chains are changing. And with retailers really not understanding their business model due to the massive shift in channel, without the type of solution that we have, it's very difficult for customers-- for our clients to make decisions about where to put their resources in these types of changes.

What we do is we produce basically this profit landscape. So we can produce a contour map that shows the peaks of profit by customer product and channel. And we can also show where the profit drains are. We use these types of profit maps to help companies understand the underlying profitable patterns that you can't see in an average profitability of 6%.

So we're using-- underneath the math here, we're using cluster analysis. We basically will look at your business through the lens of customers. We can identify your most profitable customers and your least profitable customers. We can also then do the same thing for products and for all your operations, whether that's a manufacturing line or whether it's a distribution center or whether it's a retail store or it's a service organization. We can look at the profitability below of those.

A manufacturing case study we did with the $3 billion equipment manufacturer-- we configured the business model for them in less than six weeks. We were able to identify opportunities to improve their profitability by over 50%. They selected SKU management to focus their profit improvement efforts, which represented about half of the total opportunity. And we were able to help them understand how their service level problems and their SKU perforation were having a huge impact on their profitability and gave them a solution that allowed them to create an internal hurdle rate and created a capability for them to change how they priced new SKUs and how they brought them into the company.

This generated a 14% improvement in profitability in 12 months. So we have experience across multiple industries-- manufacturing, distribution, retail services, including financial services and transportation. We are offering a six week pilot where we can go in and configure the model for a company and then quickly show them the value of the solution and then move forward with our ongoing value proposition, which is basically an annual subscription. So thank you very much.

SPEAKER: Thanks John. We'll go on to a couple of questions from the audience. Do you also share next best actions and recommendations. For example, can you increase profits by rationalizing the White House, et cetera?

JOHN WASS: Yes, we basically worked with a large company to do essentially a real time activity-based cost analysis of all the activity inside the distribution center and were able to isolate which products and how they're being picked to help them understand their overall profitability.

SPEAKER: Great, and also let's put this in perspective now with the pandemic. How does COVID 19 market disruptions affect the benefits of Profit Isle solutions?

JOHN WASS: Now more than ever, understanding where your true profits are is critical because if you're making changes or layoffs or restructuring your organization-- if you don't truly understand where you're getting your profits from, you could easily expose some of the core profitability of your company to a major change. And in our experience, something less than 15% of your customers are doing somewhere between 150 and 200% of your profits. If you expose that small group of customers to some of the broader changes that you might be anticipating, you could have a very rapid negative impact on the profitability.

SPEAKER: Thank you John.

JASON BARTON: Hello everybody. My name is, as Marcus just said, Jason Barton. I'm the Chief Commercial Officer at realtime robotics. Realtime is a startup based in Boston. We're about 40 people strong right now. And we're delighted to say that two of our founders are MIT alumni. And we continue to enjoy a very great connection with the MIT community as it stands right now. So realtime's technology can provide value across the board to robots essentially working in any environment.

However, today as a company, we focus on factory automation and autonomous vehicles. And for the speech today, I'm going to be talking about specifically factory automation. So we've all heard that the robot revolution is coming. People always talk to the automation is going to take over. But really the reality is quite different. Currently robots only account for less than 0.3% of the billion people that are employed for dynamic motor skills today. So we're barely scratching the surface. So why is this?

The problem is robots are physically very capable. But they're dumb. They exist in very rigid, structured environments doing very repetitive tasks with very little or no flexibility. Robot programming is primitive and very complex and typically accounts for 40% of the cost of deploying a work cell. And it's this complexity and cost that's prohibiting automation to really scale as it should.

So we at realtime believe we have the answer by automating this robot planning process and eliminating the complexity of robots and sensors as they're connected with automation. Our solution can reduce robot programming by 80% and total deployment costs by up to 35%.

So how do we do this? We do this by using the realtime controller and the latest advancements in machine learning and artificial intelligence. The realtime controller can plan and test millions of motions at millisecond speed, simultaneously coordinating and optimizing multiple robots in real time. If you look on the left hand side, present technology today, so the conventional way to program technology in robotics is very much like the old way of giving directions.

I could give you directions to get to the airport, but if I gave you a wrong turn or a road was closed, the directions become very quickly useless. Thankfully, today we have GPS. So to get wherever you need to go, you simply have to enter your goal location and GPS will get you there on time and will deal with changes along the way. The realtime controller is essentially the GPS for robot systems.

In this example, you see our realtime controller is controlling for robots that are very, very tightly placed together. In conventional programming, this would essentially be impossible. It would be too long to be able to program each individual motion. And you wouldn't be able to program them in a collision-free manner. With our system, all we need to do is set the start location and the goal location. And then our robots will automatically plan collision-free and get to their locations.

You can see in the left hand side there the sweat volumes that the robots are taking and how they avoid each other. We are working with over 40 global companies throughout the globe in North America, Asia, and also Europe. And here are a few examples in the automotive space and also the e-commerce space. In the second example here, you can see an example of our partnership with Siemens. We recently announced partnership where our technology will be offered as a component of their industry-leading software simulation program, Process Simulate.

With this particular customer, an automotive company in Germany, we were able to show them an 83% reduction in the programming time and cost of developing this work cell. They're now going to be rolling this technology together with Siemens out throughout their simulation group as well as throughout their factory floor. Realtime is transforming how industrial robotic applications are programmed, deployed, and executed and tackling head-on the complexity and cost that is prohibiting manufacturers to automate more frequently.

We're looking to focus our efforts on the automotive space on electronics and logistics. And we welcome opportunities and introductions to end users says end users and also system integrators in these spaces. Thank you very much for your time.

SPEAKER: Thank you Jason, great. Let's move on to some questions. Despite the growing demand for robotics, why has it been so difficult and costly for companies to bring them on or implement them?

JASON BARTON: Yeah, I think it boils down to this complexity. So if it's a large manufacturer or even small, medium sized manufacturers, everyone is keen to automate more. COVID 19 and the pandemic is really underlined this even more, I think in terms of the need to be able to have alternatives to human labor. But really the biggest stumbling block and the biggest issue is the complexity and cost of automating. So the ROI hasn't been there.

So as a company and as in many, many companies in this market are really looking to try and reduce that friction and reduce that prohibitive wall to being able to manufacture more by just making it generally more easy and more seamlessly integrated.

SPEAKER: Great. And can you talk a little more about which market holds the biggest opportunity for realtime robotics or a few of them?

JASON BARTON: Sure. Yeah, so up until now we've spent most of our time in automotive and electronics. That's been our focus to date. But again, due to the COVID 19 pandemic, and I think this was already growing this way anyway, but the logistics market fueled by e-commerce is becoming significant for everybody, I think. And we're seeing a lot of inbound interest into our kind of technology to help them automate their processes more efficiently using our vision systems to be able to analyze different parts of their operations, as well as using our technology to reduce the complexity and improve their throughput.