The
Next Decade Of The Industrial IoT
2020
AT THE ___ HEART OF EVERY MACHINE IS THE ___ BRAIN
In a three-part series, McRock Capital and Cisco Investments share their top trends in Industrial IoT for the next decade.
Content Introduction – McRock Capital PART 1: A Market Shift from Horizontal to Vertical PART 2: Edge Computing. Think Big. Go Small PART 3: Embracing AI in the Post-COVID World
McRock Capital, the premier venture capital firm dedicated to investing in the Industrial IoT, has partnered with Cisco Investments to share a vision for the top digital transformation trends that are shaping the future of every industrial market. In this e-book we explain the most valuable trends accelerating the digital transformation. Important topics like the extended reach of edge computing and the supreme competitive advantage created by Artificial Intelligence (AI) will be presented in a way that anyone leading a traditional industrial company can understand. The technologies behind the Industrial IoT are providing new tools that are changing the competitive landscape, creating growth opportunities for the digitally savvy and threatening the slow adopters regardless of incumbency. The digital transformation is also a critical part of other major global trends that are challenging humankind. As we move to correct the environmental impacts we have inflicted on the planet, the digital transformation will play a vital role in unleashing technology to help attain the sustainable resources needed to thrive on earth. The global pandemic is also accelerating the urgency for a digital transformation as remote asset monitoring is leading us towards concepts like the self-optimizing plant.
Investment Dollars have reached US$9 billion in the first half of 2020
Investment dollars into the Industrial IoT have reach an all-time high in the first half of 2020 at US$9 billion. At the halfway mark of the year, investment amounts are already double that of 2018 and 35% ahead of last year’s total. This acceleration in activity, despite the economic and logistics challenges caused by the global pandemic, is an indicator of the importance placed on industrial digital transformation projects. Robotics & automation, remote monitoring and control, digital twins and predictive analytics are high priorities for many corporate digital initiatives. As we look forward into the next decade it’s clear that a digital transformation is well underway. Many large companies have formed digital business units to lead the evolution. The foundational technologies exist and corporate resistance to external innovation in areas such as data science is lowering. The IoT will transcend the initial role of a machine data collection tool or an expensive and complicated alerting system for critical asset monitoring. Automation and autonomy, driven by machine learning and artificial intelligence, will be the end-point of the next decade of Industrial IoT. - Team McRock
TRENDS IN INDUSTRIAL IoT: PART 1
A Market Shift from Horizontal to Vertical The implementation of Internet of Things (IoT) has become widespread over the last decade. Before then, Machine to Machine (M2M) communication was the standard for machine connectivity. It started in the 1990s, when low-cost RFID tags were first placed on products to track them through the supply chain. By leveraging innovations in low-cost sensors, connectivity standards, the cloud and big data, IoT became a standard mass-market technology with applications in both consumer and enterprise settings. Many IoT startups, IT product and service vendors, and industrial solution vendors initially tried to approach Industrial IoT from a more horizontal perspective by combining IoT projects with broader enterprise digital transformation projects, in an effort to find solutions that worked across multiple industries and applications. However, over the past 5–7 years, and after several billion dollars spent on deployments, that perspective has evolved. Today, many have come to understand the wisdom of a more vertical approach for addressing specific customer pain points and use cases in an effort to deliver tailored solutions and measurable outcomes.
The advantages of a “vertical� IoT strategy Companies with a vertical strategy develop products and solutions primarily for a specific industry which may incidentally apply to adjacent industries. These companies develop deep domain expertise that can accelerate their growth, enhance the stickiness of their solutions and, over time, provide them with a competitive advantage. When analyzing the flow of investment dollars in IoT, there is strong evidence that vertical Industrial IoT startups are attracting more investors and acquirers. This trend looks to continue for the near term, given the ability of vertical Industrial IoT startups to build a strong line of defense with subject matter expertise, proprietary data, partnerships and products/services that address a real market gap in their respective industry. Let’s look at some of the investment trends for vertical* and horizontal** Industrial IoT startups over the past five years. The total funding for horizontal IoT solutions started strong in 2014 with $787 million*** invested across various startups.
* Includes IoT companies in Manufacturing, Mining, Oil & Gas, Construction, Agriculture, Logistics and Smart Cities ** Includes Sensor Platforms / Stacks for prototyping/building apps along with software platforms for Network Management, Device Management, Data Management, Analytics and Network Service Providers/Enablers, specifically for IoT applications *** Source: Pitchbook, Tracxn
Over the next three years, dollars invested in horizontal IoT solutions increased steadily, but experienced a sharp decline in 2018 with only $658 million in investments – a decline of 61% compared to the previous year. Total investments in 2019 however are estimated to have hit $1 billion, with few one-off big rounds of late-stage financing, including Samsara ($300 million), SparkCognition ($100 million), C3.ai ($113 million), and RootCloud ($217 million).
There is strong evidence that vertical Industrial IoT startups attract more investors and acquirers.
By contrast, vertical IoT solutions had a humble beginning in 2014 with about $349 million in total investments, but that number has grown strongly at a 75% CAGR to an astonishing $5.7 billion in 2019, i.e. 5.7x more capital raised than horizontal IoT startups in the same year. When it comes to acquisitions, horizontal IoT solutions companies witnessed higher acquisition activity over the past few years by virtue of being developed ahead of vertical solutions companies. Cisco’s acquisition of Jasper in 2016 is one such example. However, during 2018-19, M&A activity in the vertical IoT space picked up steam. The acquisitions of Pure Technologies by Xylem, Trafficware by Cubic, and Cisco Investments portfolio companies Relayr by Munich Re and Bit Stew by GE are notable examples. Why are investors and acquirers more interested in vertical IoT startups, and what does it take to build a leading company in this space? We have some thoughts on that.
61% Decline in horizontal IoT investment dollars from 2017 to 2018
$1B Invested in horizontal IoT startups in 2019
5.7x More capital raised by vertical IoT startups vs. horizontal
Key success drivers for vertical IoT companies Unlike fixed and mobile Internet, IoT is not standards-driven but applications-driven, with different user and buyer personas for any given vertical. Building a vertical IoT solution requires focus on solving a specific problem for specific industries. We see many founders trying to build a vertically focused IoT company with a technologycentric approach rather than a customer-centric approach By focusing solely on a customer’s specific needs, a startup can narrow the list of assets to be connected, the types of sensors to be deployed, and/or the data sources to be considered in an effort to train its algorithms most efficiently.
For example, Invixium, a McRock portfolio company, zeroed in on a work force management use case utilizing its cloud platform, analytics and biometrics-based access control technology. Using a very specific set of applications, the company managed to significantly influences how organizations within that vertical can manage and track their people. Today, Invixium serves more than 300 customers in more than 60 countries and has achieved significant commercial scale. A vertical solution usually comes with a well-defined set of supported devices, sensors and software from multiple vendors.
Due to lack of device and protocol standardization in IoT, the addition of more software or devices to an existing network may become time consuming and resource intensive. Vertical IoT startups therefore, need to proactively mitigate scalability and lock in issues to succeed. Further, for defensibility of the business, it is key to ensure sustained access to IoT data through proprietary sensors or long-term, contractually secured southbound connections.
‌a vertical IoT company should continue to expand TAM over time by extending into more use cases in their focus verticals
Due to lack of device and protocol standardization in IoT, the addition of more software or devices to an existing network may become time consuming and resource intensive. Vertical IoT startups therefore, need to proactively mitigate scalability and lock in issues to succeed. Further, for defensibility of the business, it is key to ensure sustained access to IoT data through proprietary sensors or longterm, contractually secured southbound connections.
Kespry, a Cisco Investments portfolio company that provides dronebased aerial intelligence, first created an autopilot solution for aerial data capture. This autopilot solution had no hardware casing or airframe mounts around it for industrial use case applications. So they built a proprietary drone suitable for industrial data capture. They then went on to build data infrastructure in the cloud to ingest image and video data from the drone to create machine learning-based insights for industrial customers. Building a vertical IoT Startup in the Industrial space demands a deep understanding of that vertical’s value chain. As most IoT projects remain heavily inflenced and funded by operational technology (OT) buyers, a vertical IoT play allows startups to tailor their solution’s value proposition messaging to the key pain points of OT buyers in that vertical. Further, as initial TAM may not be as large as that targeted by a horizontal company, a vertical IoT company should continue to expand TAM over time by extending into more use cases in their focus verticals.
Veniam, a Cisco Investments portfolio company, demonstrates how to do this. The company started as an intelligent mesh networking solution for vehicles and has gradually expanded to become a platform that enables applications and services to access a vehicle’s data, communications and computing resources with the right QoS and security.
We cannot emphasize enough the importance of domain knowledge for a successful vertical IoT Play Selling these products requires trust and relationships within the industry. Teams that combine technical and subject matter expertise are better able to use outside-the-box thinking to model the domain and drive innovation because they have a thorough understanding of what the box actually is. McRock Capital invested in one such company, Decisive Farming. The founding team leveraged its deep domain expertise in agriculture, process management and analytics to develop an integrated platform that became the operating system of their customers’ farms. By knowing the various stakeholders and addressing the pain-points across the value chain, Decisive Farming expanded its platform’s offering to three core farm functions: Farm Management – to improve performance, Precision Agronomy – to increase yield, and Crop Marketing – to grow farmer revenue. As strong as the market potential is for vertical Industrial IoT startups, there is still a strong upside for those startups taking a horizontal approach.
Key success drivers for horizontal IoT companies Unlike fixed Can Industrial IoT startups be purely technologydriven as opposed to customerdriven? The answer is yes, but it requires that they target use cases that occur across verticals, such as connectivity, security and data infrastructure, or that they build an incredibly strong technical team which is capable of architecting solutions that are useful in multiple industries. Examples of this include mnubo, a McRock portfolio company which was acquired by Aspentech, and Sentryo, which was acquired by Cisco. If a startup decides to follow a horizontal approach, there are two ways to differentiate: one is to focus on the democratization of the base technology, and the other is to create a strong community around consumers of that technology. To execute on the latter, horizontal IoT startups need to have robust northbound APIs and logic-building platform capabilities to differentiate themselves and allow adoption across verticals.
IoT is a convergence of markets, sub-markets and ecosystems. This fact makes it di􀋽 cult for horizontal IoT startups to appeal to the multitudes of buyers and value chain stakeholders when going it alone. It is imperative for horizontal IoT startups to partner with other horizontal vendors – in different parts of the tech stack – as well as vertical IoT companies, to deliver holistic solutions to customers. This is also an area where strong incumbents such as Rockwell, Schneider Electric, and Honeywell have deeply embedded solutions, so startups need to find a specific capability to compete with incumbents and capture value. In terms of untapped opportunities, we see immense potential for horizontal IoT startups to innovate in areas like security, wireless connectivity (5G, Wi-Fi 6 and CBRS; read more here), edge computing and analytics.
‌horizontal IoT startups need to have robust northbound APIs and logicbuilding platform capabilities to differentiate themselves and allow adoption across verticals
What’s next for vertical and horizontal IoT companies? The past decade was more about building the blocks of Industrial IoT at a time when the horizontal-approach strategy made more sense. At that stage, IoT was a technology looking for a market. Today, practically every industrial market looks to IoT for customer-centric solutions. Going forward, startups that take a vertical approach to these solutions will continue to attract a growing share of customers, investors and acquirers. For world-class IoT companies focusing on horizontal solutions, areas like connectivity, security and data management remain attractive and will continue to offer scaling opportunities. However, they are not the end goals, but rather the foundation of a path toward something greater.
Interested in learning more about McRock Capital & Cisco Investments?
If you are an IoT startup and would like to partner with us to build the nextgeneration of solutions, reach out to McRock Capital and Cisco Investments.
TRENDS IN INDUSTRIAL IoT: PART 2
Edge Computing Think Big. Go Small.
Neither a startup nor technology could disrupt our world quite in the same way as the pandemic currently facing humanity. “Going small” has never felt so real as we hunker down in our lives and businesses. Suddenly, the need to decentralize computing and enable more nimble processing at the source of data spans all industries. Scientists at the University of Massachusetts, Amherst, have recently developed a portable surveillance device that can monitor spread of respiratory illnesses and flu trends using machine learning models. Their edge computing platform, FluSense¹, analyzes data on coughing sounds and crowd size in real time and could help track the spread of COVID-19. Such edge computing systems that can provide greater agility in processing real-time data, insights, and automation at the source will emerge as a big trend as we seek to build new bridges between humans and things.
1 FluSense: https://www.umass.edu/gateway/feature/flusense
By 2025, Gartner predicts 75% of all enterprise-generated data will be created and processed outside a traditional centralized data center or cloud
By 2025, the research firm Gartner predicts 75% of all enterprise-generated data will be created and processed outside a traditional centralized data center or cloud, compared to 10% today. In our Part 1 article, we emphasized how applications, not standards, are driving the IoT market with a multitude of use cases and a wide array of personas for any given vertical. Across today’s IoT landscape, asset and application sprawl, harsh industrial environments, limited compute per node, security vulnerabilities, and changing network conditions persist as challenges, paving the way for dedicated edge computing infrastructure to unlock real value. For sake of simplicity, we define edge computing as use-case driven placement of computing hardware and software closer to the source of data.
Contextualizing the Edge Think of the IoT landscape as a four-player chessboard with varying degrees of “coopetition” among the four key incumbent vendors: IT, Industrial OEM, Cloud, and Telecom, each viewing the edge computing market through a different lens.
IT
Industrial OEM
Edge means customer onsite IT infrastructure, e.g. a server rack in a data center or an access point in a warehouse.
Edge compute deployments mean a video camera or a robotic arm on the factory floor or a connected fleet of vehicles.
Cloud Everything outside the core cloud is edge.
Telecom All deployments from the cell sites to the core span the edge.
Edge Computing Investor Interest On the Rise
According to the startup tracking platform Tracxn, VC investors funded approximately $500 million in edge computing startups from 2015 to early 2019, a modest percentage of the overall investment in IoT. However, from H2-2019 to Q1-2020, several large financings transpired, including Hailo Technologies, an Israel-based AI processor solution for edge devices, which raised $60 million in Series B funding, and edge AIfocusedchip makers Kneron and Untether AI secured $40 million funding in January 2020 and $20 million funding in November 2019 respectively. Swim.AI, the data processing and edge analytics software company, raised $18.4 million in late 2019 and Vapor IO, the micro data center and edge computing platform, raised $90 million in Series C funding in January 2020. Edge computing platform Aetheros secured $15 million in Series A funding in November 2019, while edge intelligence software startup Foghorn raised $30 million in Series C funding in August 2019. Acquisitions have also spiked over the last 12
months. For example, Equinix acquired Packet for $335 million in March 2020 with the data center and colocation company planning to leverage Packet’s technology to accelerate the development and delivery of its interconnected edge services. Earlier this year, Apple acquired Xnor.ai in a deal expected to boost Apple’s on-device AI capabilities and competitive advantages in computer vision. Also, Siemens acquired Pixeom in October 2019 to strengthen its Industrial Edge portfolio by adding software components for Edge runtime and device management. The edge computing market (spanning hardware, software, and services) is expected to grow from $2.8 billion in 2019 to $9.0 billion by 2024, at a CAGR of 26.5%, according to MarketsandMarkets². With growing maturity of edge platforms and the concurrent adoption of complementary technologies such as 5G and machine learning, investments in edge computing are expected to increase.
We expect edge computing technology to evolve from independent silos to being more “cloud-like� in this decade, progressing through three cycles
Evolution of Edge Computing Over the Next Decade We expect edge computing technology to evolve from independent siloes to being more “cloud-like” in this decade, progressing through three cycles.
EDGE 1.0
COMPUTE AT THE EDGE For the last few decades, we’ve observed traditional light footprint computing at the edge, predominantly for capture of telemetry information. Market watchers expect the trend to continue as the industry provisions more edge computing closer to remote users and specific data sources, though future growth in adoption of this first edge iteration will likely be limited.
EDGE 1.5
EDGE AS EXTENSION OF CORE We have also witnessed growing adoption of ruggedized servers and IoT gateways with a software control plane spanning the core IT. By combining ruggedized hardware with a software platform, customers would be better equipped to run applications at the edge.
EDGE 2.0
EDGE CLOUD This is edge as part of a seamless cloud architecture. Edge cloud extends some capabilities of the cloud (including but not limited to storage, computing, network, AI, and security) to edge nodes, via a tight coupling of computing and networking hardware. Software orchestrates this fine symphony of inter-node and cloud communication, as in the case of Cisco’s Edge Intelligence software combined with Cisco’s IoT networking portfolio. The days of edge cloud are still early, and all forms of edge computing are likely to co-exist for the next couple of years.
Opportunities We See at the Edge Edge computing will drive a use case-dependent evolution of the tech stack – requiring new hardware form factors and software capabilities to address scale and complexity of distributed infrastructure. We believe that several opportunities exist, both on infrastructure and applications fronts.
Edge Infrastructure Opportunities – “Edge as a service” platforms Colocation of massive data centers close to large-scale, distributed IoT endpoints is neither a practical solution nor does it address the networking challenges of today’s edge computing implementations. Instead, a new paradigm of software-defined networking is primed to balance the distributed nature of IoT endpoints-driven data generation while addressing the demand for low latency among edge use cases. Opportunities will proliferate for edge computing and networking platforms that enable this fine orchestration of data and workloads across various nodes with container support to run distributed applications. Startups like Pixeom, ClearBlade, and Vapor IO play in this segment. Data Management and Streaming The scalability, latency, and reliability demands of modern data streaming IoT applications (e.g. live video streams) running on distribution edge infrastructure are disrupting the edge cloud stack today. Controlling such a high rate of data flow and consequently a high rate of state change using the traditional, stateless data cache and streaming models has proven incredibly difficult. Edge-native distributed databases and pub-sub platforms for real-time stateful apps will prosper with their ability to manage challenging IoT data streaming conditions, such as unreliable WAN, lossy time keeping, and lack of consensus. Acting on this thesis in 2016, Cisco Investments invested in PubNub, a data streaming network for modern, real-time IoT and communications applications. Startups like Swim.ai and Macrometa illustrate complementary edge-native database and data cache functionalities.
Edge Cloud for 5G 5G and edge computing are two inherently linked technology paradigms for delivering an enriched application experience. Edge cloud supports 5G adoption with local, rather than regional, compute resources that can address 5G’s high-bandwidth, low-latency (1ms) requirements. While network functions have traditionally run on purpose-built appliances, general-purpose edge compute and edge clouds are deploying virtualized network infrastructures for 5G to replace many dedicated hardware-based elements with virtual network functions (VNFs), especially for packet forwarding and security. The combination of network function virtualization (NFV) and edge computing makes it easier than ever to manage the lifecycle and configuration of these new VNFs and helps 5G deliver on its promise of enhanced application performance.
By localizing applications at the edge closer to end users, both network transit latency and reliability improve, driving further adoption of technologies, such as industrial robotics and drones, vehicle-to-everything (V2X) communication, AR/VR infotainment, autonomous vehicles, and associated business models.
Hardware acceleration with ML focused chips The advent of real-time applications in IoT leveraging machine learning has made it clear that inference needs to happen at the edge, rather than in the cloud. Software innovation is not enough to deliver the required results for such use cases with the current breed of general-purpose hardware. Rather, we need new computing hardware that meets IoT edge characteristics, such as small form factor and low power consumption, while delivering more robust compute. We are seeing growing adoption of ondevice chipsets optimized for faster, more secure, use case-specific inference developed by Intel, Google, and startups such as Mythic.ai and Untether.ai.
Edge Application Opportunities Edge devices collect vast amounts of information for real-time processing, which leads to the biggest opportunities for edge applications within three key categories: Sense, Automate, and Analyze. Edge as a sensor By 2025, the total installed base of IoT connected devices is projected to reach 75 billion worldwide³, the largest proportion of which will comprise inexpensive sensors deployed to generate and collect real-world data about ourselves, our machines, and our environment. These edge sensor networks form the lifeblood of real-time applications and other significant opportunities for companies that are creating the foundation for edge computing through proprietary sensor-based data moats. Miovision, a McRock Capital portfolio company, is one such company empowering “smart cities” by accumulating one of the world’s largest video traffic data repositories, counting more than 9 billion vehicles and 850 million cyclists and pedestrians. Through sophisticated sensors and AI, Miovision brings visual data collection roadside to help cities sense and understand what’s happening at any intersection in real-time.
Smart edge with ML-driven automation Robots as well as self-driving cars and trucks feature use case-driven, fullstack applications that leverage edge computing, machine learning, and other modern technologies. For these reasons and more, we’re seeing massive value-creation opportunities with smart machines operating at the confluence of edge computing and machine learning. For example, AImotive, a Cisco Investments portfolio company, provides full-stack, AI-based autonomous driving solution for cars, including edge computing hardware for AI inference in the car and automated driving software. Almotive’s edge compute platform gives the car’s smart sensors and compute clusters real-time inference and autonomy. Even today, in the fight against COVID-19, we’re witnessing smart edge applications rise to the occasion, including the use of robots to ensure social distancing, especially in high-risk areas. Across several provinces in China, hospitals are using robots to deliver food to patients, to help clean the facilities, and to avoid unwanted contact by navigating patients between hospital zones. Edge driving real-time analytics Unlike traditional analytics models that depend on centralized data warehouses and lakes, edge analytics models collect, process, and analyze data at the edge with an emphasis on speed and decentralization. Over time, we expect use case and device-optimized machine learning models running at the edge to provide more value. Consider startups such as FogHorn and Litmus Automation, which have employed the ML approach to offer closed loop edge analytics solutions for industrial customers.
Scaling Edge Startups Despite the potential for unlocking massive value, edge computing has yet to result in many scaled up startups or large-ticket acquisition deals to date. We have some thoughts on this. While still in a nascent stage, the edge computing market continues to carry high customer risk aversion, namely for the extreme diversity that characterizes the different systems and networks across locations. Let’s dive deeper into specific challenges and success factors.
Edge Infrastructure Infrastructure platforms for edge, such as Edge as a Service and Edge Cloud for 5G, need to provide platform features, such as low- or no-touch remote provisioning and management, security, scalability across locations and network conditions, as well as out-of-box integrations with incumbent IT and OT solutions. Whereas startups may struggle to bring all these features at the outset and serve as the chosen platform for edge implementations, incumbent platforms in computing, network management, and automation markets are much better poised for success in this regard. Startups offering use-case specific infrastructure components and tools will complement the broad offerings of edge platforms. One such opportunity lies in data management for tools to offer ingestion, transformation, streaming, synchronization across edge nodes and cloud, and governance features to complement the edge infrastructure platforms. Also, startups will find opportunities in providing technologies that support machine learning inference at the edge, including software optimization to run ML models on low-power, low-compute edge devices, such as Xnor.ai, and hardware acceleration with ML-focused chips.
Edge Applications The following strategies can help applications startups win at the edge: 1. Tap into existing budget with measurable ROI: Edge applications should target a broader story of pervasive connectivity, sensors, and business-critical use cases where the ROI is clear to demonstrate. As part of digital transformation or Industry 4.0 initiatives, most enterprises have existing or emerging budget allocations for AI, 5G, automation, remote monitoring, and predictive maintenance, all of which typically have an edge component. 2.
Continuously learn and improve: As more data is generated and captured at the edge, new opportunities will emerge for expansion into greater use cases and those deeply entrenched within customers’ accounts.
3.
Leverage the cloud: It’s not edge versus cloud; it’s edge and cloud as synergistic technologies complementing one another. Edge startups should partner with leading cloud players, wherein edge startups bring domain-specific solutions to the partnership, while cloud vendors bring top-down sales motion in addition to cloud compute and storage.
4.
Optimize spend: Startups should optimize their infrastructure use, accounting for the following: 1. Compute and storage: Much of data collected at the edge is often disposable as it does not reflect a useful or critical change in the operation or condition being monitored. Therefore, startups should be smart about the way they sample, store and infer edge data. 2. Network bandwidth: Much edge-generated data is only valuable at the edge and is at times only useful for a short period of time (few hundreds of milliseconds).
5.
Partner with incumbent IT and OT platforms: In a nascent market, it is tougher for startups to scale customer accounts beyond trials. Bestcase scenarios are partnership models in which the startups’ application is part of a bigger project with pull-through from incumbent IT and OT platforms.
TRENDS IN INDUSTRIAL IoT: PART 3
Embracing AI in the Post-COVID World The coronavirus (COVID-19) pandemic is the black swan event of a lifetime. It has shocked the world, upended industries, nearly broken several countries’ health infrastructure, and brought economies to a grinding halt. Travel restrictions and lockdowns have subdued demand across transportation, energy, and retail, among other verticals, while stalled production in China and Europe has disrupted supply chains around the world and increased uncertainty about the future of the global economy. Following the pandemic, many industries will, without a doubt, look entirely different from what they were before. New challenges and opportunities will surface, and questions previously overlooked will now demand answers. Artificial Intelligence (AI), a technology which was in the nascent stages of adoption in industrial verticals, will lay the foundation for significant investments in digitization and automation in the post-COVID world.
“Following the pandemic, many industries will, without a doubt, look entirely different from what they were before.�
Clearpath Robotics
An AI Framework to Optimize Assets, Improve Processes, and Benefit Humans Over the past several years, AI combined with IoT has led the charge in driving greater efficiencies across many industries. And yet, despite the significant promise of the technology, many organizations have not developed strategies and critical skills to fully realize the value of AI at scale. Many business leaders who prioritized solving other problems with a shortterm results horizon failed to make long-term investments in preparing for the impending automation revolution. The disruption of the pandemic, however, has forced organizations to rethink their digital transformation approach and accelerate the adoption of AI applications. As great uncertainty around the economy’s recovery and available capital for growth and expansion looms, businesses will increasingly look to AI-led automation to drive more value from their existing applications.
AI also has the potential to create new business models and applications that will profoundly change the way people live and work post-COVID. McRock Capital and Cisco Investments created the below framework to illustrate how AI can optimize assets, improve processes, and enhance human capabilities through a mix of existing and emerging applications.
“AI-enabled software tolls have proven their value in augmenting human capabilities to improve productivity and optimize workflows
AI Supercharging Existing Applications The Augmented Industrial Worker Improvise and adapt to overcome short-term disruptions The advent of digital technology has already shifted much of the physical work to digital. AI-based technologies had become quite pervasive prior to the COVID-19 pandemic; the disruption caused by such event, however, has further expedited the adoption of AI in the workplace. Keeping frontline workers safe: As factories, warehouses, and other work sites re-open, many are looking to use AI-based solutions to safeguard workers’ health. Some of the safety applications include alerting management and workers in real time to potential safety hazards as well as ensuring employees are wearing face masks and other protective equipment, adhering to social distancing guidelines, and receiving skin temperature checks through advanced thermal imagery. Accomplishing more with fewer resources through AI-driven analytics: While there are arguments to be made around AI’s negative impact on the global job market, the pandemic caused a decrease in human workforce overnight and made AI more appealing. AI-enabled software tools have proven their value in augmenting human capabilities to improve productivity and optimize work ows. Specialized jobs, in particular, bene t from machine learning algorithms that identify trends and insights in vast reams of data and enable faster decision-making. One case in point, ThoughtTrace, a McRock portfolio company, has developed an AI-powered Document Intelligence platform that delivers rich, contextual insights for complex documents and contracts. Using advanced Natural Language Processing (NLP) techniques, the solution maintains a human-like understanding of unstructured document data, giving customers an enhanced view of their repositories and extensive insight into their complex contract data to drive actionable business value.
A shift in workforce foundation As a remote and virtual work model reshapes the “new normal,� technologies such as AI will become the major driving force to reimagine the future of work across different industries. Industrial worker collaboration: Many experts expect a major rise in adoption of AI-enabled connected technologies for the industrial workforce, some of which is shifting permanently to remote-based operations. Connected Worker and Augmented Reality solutions that enable remote operations and collaboration will benefit from this trend. Upskill, a Cisco Investments portfolio company, is pioneering this movement through its software platform by delivering enterprise apps for industrial workers on smart glasses, smartphones, tablets, and augmented reality devices. Their solution enables workers to receive complex assembly instructions, perform tasks faster, make fewer mistakes, avoid accidents, get remote assistance, and update systems of record in real time, all of which enables industrial customers to do more with constrained human capital.
Powering actionable knowledgesharing and creation remotely within the workforce: The amount of knowledge that needs to be managed and distributed increases exponentially as organizations grow. AI allows the system to automatically surface the right information at the right time to people in need. The use of AI helps employees discover useful resources they did not even know existed and connect people to the right subject matter experts in the company. As COVID-19 continues to alter the workforce structure, effective knowledge sharing within or across companies will assume a significant role in creating competitive advantages and improving positive business outcomes for industrial organizations.
Using a data-driven approach for workforce segmentation: According to Forbes, one of the key trends shaping the human capital market is the perspective to treat individual employees uniquely, particularly with respect to their workplace contributions. The post-COVID remote work era calls for assessing employees’ skills, strengths, and contributions and empowering them with a tailored approach (remotely), which AI-driven classification and segmentation tools could address. Organizations have started using NLP and predictive modeling to look at internal and external data on critical and growth skills, talent movement, and role adjacencies to inform talent investments. Leveraging the power of AI analytics will shed light into organization’s hidden potential and drive meaningful insights for talent retainment, reskilling, and upskilling of the workforce, as well as cultivating a culture of diversity and inclusion.
“The post-COVID remote work era calls for assessing employees’ skills, strengths, and contributions and empowering them with a tailored approach (remotely), which AI-driven classification and segmentation tools could address.” Extended Health and Safety of Industrial Assets Predictive/prescriptive maintenance: Advances in AI and data analytics continue to push prescriptive maintenance forward in generating predictions and recommendations regarding asset failures. New players continue to emerge in different parts of the stack, from sensors to platforms and applications, with successful vendors being able to convert customer data as-is from the source into actionable insights from Day One.
Praemo, a McRock portfolio company, is among those whose mission is to transform asset maintenance through AI analytics. The company combines industry expertise with advanced machine learning algorithms to automatically create out-of-the-box actionable recommendations on potential equipment failures and process optimization. As the definition of assets expands, the topic of asset data itself has become more than just about the technology. What will set successful businesses apart in a post-COVID world will be their ability to manage data as an asset across the whole organization. Integrating, correlating, and contextualizing information from different systems is key to delivering insights and creating competitive advantages for enterprises. In addition to the big players in the asset optimization space such as Aspentech, emerging startups have begun to offer unified data repositories that allow consistent management of data at scale. Physical security: The challenge of protecting facilities and assets from afar has become even more daunting as the majority of the workforce shifts to remote operation. AI technologies such as computer vision, pattern matching, and predictive analytics continue to advance, with governments, enterprises, and individuals applying this technology to video surveillance and physical access control. In addition to cameras with limited field of vision, organizations have started using drones to guard perimeters of critical production facilities.
“Integrating, correlating, and contextualizing information from different systems is key to delivering insights and creating competitive advantages for enterprises.�
Best Practices for Maintaining a Strong Supply Chain Product quality control/inspection: COVID-19 has created a severe impact on the supply chain, causing industrial companies, especially those in manufacturing, to face new constraints in resources while maintaining product quality to mitigate losses. Such pressure will accelerate the adoption of AI in automated quality inspection. Computer vision, for instance, uses cameras to scan the device or part under test for catastrophic failure and/or quality defects. Semisupervised machine learning then classifies images into various failure classes. AI is also used to analyze other sensor data in time series format for ensuring quality of manufacturing output. AI analytics for demand forecasting: Demand planning serves as the starting point for many other activities across the supply chain, such as shipping, warehousing,
pricing, and especially, supply planning. Demand forecasting systems have been around for decades, but they are vulnerable to large-scale disruption, such as the COVID-19 pandemic. As the requirement for high accuracy predictions becomes more critical than ever, AI solutions can help upgrade statistical models according to the current reality. For instance, AI techniques like NLP allow the systems to add and analyze new data points, such as human behavior, including those from media platforms and conversations, to detect and predict people’s preferences, choices, or sentiment. AI analytics for logistic operations visibility: Building a resilient supply chain in an uncertain post-COVID world requires greater visibility into every part of logistics operations.
Solutions that integrate various sources of data such as weather, social media, border crossing times based on time of day and day of the week, inventory level, and other relevant information can provide advanced warning to a facility when a shipment is delayed, for example. The system can then use AI analytics to lower the risk of supply chain disruptions by continuously recommending resolutions that help avoid bottlenecks and notify customers regarding potential changes or delays.
“The applications of Reinforcement Learning will unburden the human operator from having to pre-program accurate behaviors, allowing organizations to implement flexible and responsive robotics and industrial automation systems adaptively in a broader range of unstructured and unknown environments and use cases.�
Clearpath Robotics
Emerging Applications: What’s on the Horizon? Innovations in AI algorithms and frameworks are spawning new applications of AI. One such promising technique is reinforcement learning (RL), a deep learning technique that enables an agent to learn in an interactive environment by trial and error using feedback from its own actions and experiences. The goal of RL is to find a suitable action model that maximizes the total cumulative reward of the agent. Over the next decade, reinforcement learning is expected to fuel the growth of emerging realworld AI applications. Moving from industrial automation to autonomy: AI will catalyze paradigm shift from connecting assets & processes and performing descriptive and predictive analyses to adoption of adaptive, selfoptimizing systems in the post-COVID world. RL is an ideal solution to use when an organization is grappling with real-world challenges in automating existing processes. RL enables a robot to autonomously discover an optimal behavior through trial-and-error interactions with its environment, such as picking an object from a box and placing it in a container a la FANUC. Even if it fails, it learns from the experience and improves subsequently. Other applications of RL in industrial world include process and machine calibration, intelligent control systems design for traffic lights in smart cities and HVAC systems in energy-efficient buildings, and autonomous systems development for mining and drilling operations. The applications of RL will unburden the human operator from having to pre-program accurate behaviors, allowing organizations to implement flexible and responsive automation and autonomous systems in a broad range of unstructured and unknown environments and use cases. Adaptive inventory management and delivery optimization: Typically, different actors within a supply chain, namely suppliers, manufacturers, and distributors, have different inventory policies. Reinforcement learning algorithms can be deployed to reduce transit time for stocking as well as retrieving products in the warehouse. RL can even optimize deliveries by serving several customers with one vehicle.
“Successful AI companies are those who develop effective frameworks that enable them to proactively keep up with data advantages through constant analysis of data distribution, impact of data on product finetuning, as well as the trade-off between volume and quality.�
Scaling AI Startups With the potential to unlock massive value, the AI market is teeming with startups as well as incumbent technology vendors’ solutions. The market, however, lacks a robust framework for customers to compare and choose AI solutions most relevant to their needs, which makes it challenging for startups to stand out in this burgeoning, noisy market. We have some thoughts on how AI startups can compete and scale.
Product and Talent Large and proprietary data training sets coupled with a thoughtful data strategy: Data is the new oil that will help AI systems bolster their efficiency over time. However, long-term competitive moats won’t be created through data collection alone. Successful AI companies are those who develop effective frameworks that enable them to proactively keep up with data advantages through constant analysis of data distribution, impact of data on product finetuning, as well as the trade-off between volume and quality. A consistent pool of talent around applied AI: While there are conflicting reports about the number of people who actually make up the AI talent pool (between 10,000 and 300,000 globally), experts agree that “AI builders” are the most sought-after professionals when it comes to companies striving to fill their AI skills gap. Winners in this space are those who can recruit and retain AI talent through a culture of growth, not just compensation. In a competitive market for AI talent, a diverse and inclusive workplace is key to attracting exceptional talent and broadening an organization’s perspective.
Domain knowledge that provides deep insights into opportunities within a sector: Companies with combined expertise in data science and machine learning as well as industrial domain can help customers solve critical business problems and build a good rapport through customer support. Intelligent customers are especially adept at catching lack of information, misdirection, or ineffective communication. An organization's in-depth knowledge of every aspect it handles will not only separate it from its competition, it will also impress upon the customer that the said organization is best suited to tackle the problem at hand. Know and reign in direct costs: AI companies are believed to have lower gross margins than other SaaS businesses, due to heavy cloud infrastructure usage and continuous humanin-the-loop requirements. Economies of scale can only increase the margins slightly. Smart approaches around efficiently using training data and compute as well as unique pricing models, in which customers share cloud and human support costs, could help lower the expenses. AI-enabled hardware–understand your supply chain: For those building robotics and other smart hardware solutions, the challenge of building a competitive physical product adds to the complexity of scaling. We believe the key to succeed is to focus on building only the key new features that deliver the proprietary value and leverage off-the-shelf components. Understanding the pros and cons of various options for the supply chain can help lower potential risks related to IP defensibility, working capital, and production disruptions. Take advantage of services: For AI startups that provide complex enterprise solutions, having a services component is critical on multiple fronts. Deploying a well-established service strategy enables new AI vendors to learn about the real customers’ pain points early on, gather relevant data for model development and finetuning, enhance their product’s features, and ensure stickiness of the product down the road.
“As most executives and corporate buyers are not typically fluent in AI, startups need to be 'bilingual' and provide context on why a certain technology can help drive a business forward—by impacting revenues, efficiency, and customer service.”
Go-To-Market Focus on high-value problems: To gain the attention of high-level decision makers, AI solutions must address critical problems that affect the organization’s greater performance and financial outcome. Identifying high-value use cases is the key to cultivating and converting a valuable and sizable customer base. Convert technology into business transactions: As most executives and corporate buyers are not typically fluent in AI, startups need to be “bilingual” and provide context on why a certain technology can help drive a business forward—by impacting revenues, efficiency, and customer service. It is critical that startups translate between technical and commercial languages.
Be the Yoda: Selling requires time to identify and engage with stakeholders; assess the current business situation; ask questions to address critical problems; evaluate the implications of those problems across multiple departments; and build a consensus around the implications. Startups need to build rapport with key individuals in the organization and make sure the users and customers see them as trusted advisors for their business problems. Finding internal champions will help push progress forward.
Aim to delight: Make it simple to implement a PoC or pilot. AI startups should not burden customers with acquiring and cleaning data just to set up a pilot. Go above and beyond to make sure you deliver whatever they need. Prioritize, then customize: Enterprise sales, especially in the industrial market, entails a long process with the typical sales cycle ranging from nine to 18 months. Many AI startups run out of money while allocating resources to pursue less than profitable sales opportunities. Build a playbook to optimize the customer qualification process and drive growth with constrained resources. Learn to say “no” to the “not-so-right” use cases and customers.
Let’s Build the Next-Generation Intelligent Solutions Together If you are an AI startup and would like to partner with us to build the next-generation intelligent solutions for the physical world, reach out to McRock Capital and Cisco Investments.