Autonomous Disinfection of Public Spaces

COVID-19 has brought to light the importance of disinfecting public space for global safety. Sol Robotics proposes a low-cost autonomous robotic system that is capable of identifying and disinfecting high-contact surfaces within public settings. The proposed system would include an AI driven autonomous mobile robotic platform equipped with an actuated arm that supports an array of sensors and a pathogen-killing ultraviolet (UV) light source. We believe the proposed technology will have the potential to reduce the pathogen levels in hospitals, schools, airports, sporting arenas, and offices, along with thousands of other heavily trafficked public spaces. This would result in improved safety for millions, if not eventually billions of people.

One of the primary means of transmission for COVID-19 and other pathogens is touching high-contact surfaces within public spaces, such as hospitals, schools, and airports. Disinfecting these surfaces is possible to do by hand, but is labor intensive and primarily done using solvent-based cleaning products that are not always readily available, either due to supply shortages or cost. This is the problem that Sol Robotics aims to solve. We will offer an autonomous mobile robotic system that can intelligently identify and disinfect high-contact surfaces in order to offset manual labor costs and do so using sustainable and economical ultraviolet light instead of expendable cleaning products.

Sol Robotics will offer a low-cost robot that is capable of identifying and disinfecting high-contact surfaces within public settings. The proposed system would include an AI driven autonomous mobile robotic platform equipped with an actuated arm. Mounted to the actuated arm would be a commercial grade short-wavelength ultraviolet (i.e. UV-C) light wand with an energy output proven capable of killing pathogens. Also mounted to the mobile platform will be an array of sensors, including infrared camera(s) and laser light sensors (i.e. LIDAR) for detecting objects. High-contact surfaces would be identified via distinct thermal gradients that would be indicative of recent biological presences (e.g. a warm chair), along with AI-trained object recognition of surfaces deemed as likely contaminated (e.g. tables, handrails).

Sol Robotics is looking to help communities that are impacted disproportionately by COVID-19 as part of our go-to-market strategy. This includes offering a low-cost solution to help provide safe, infection-free public spaces within these communities’ schools, hospitals and businesses. We are currently interviewing hotel and hospital workers to better understand what type of product features are necessary for these industries, as well as what price points are acceptable. We are also speaking with local Arizona government officials in the Maricopa County Department of Public Health to better understand how the local government can help in the deployment of our solution.

Sol Robotics aims to fight the spread of COVID-19 and other pathogens, wherever they exist. While our go-to-market strategy targets domestic businesses in the hospitality and medical industries, we hope to eventually build toward having a fleet that can be deployed domestically and internationally at the request of governments that are battling pathogen outbreaks.

Sol Robotics is developing an AI driven robotic solution to disinfect public spaces with the intention of slowing the spread of COVID-19 and other pathogens. We aim to help low income communities by involving local government bodies to help support our product. For these reasons, we believe that our goals as a company are well aligned with those of the Challenge.

Non-targeted disinfection using UV-C lamps has been proven to be an effective method of disinfecting unoccupied spaces. It has been used in several commercially successful robotic systems, such as Blue Ocean Robotics’ UVD Robot and Xenex’s LightStrike to disinfect hospital rooms and surgical theatres. It is important to understand, however, that these competitor’s disinfection robots are not safe to use around people. This is because they use non-targeted UV-C that can burn eyes and skin if exposed to it. This is the key differentiator between Sol Robotics’ product offering and our competitors. Our disinfecting robot is safe for use around people. Our product can be used in public spaces, such as hospital lobbies/cafeterias, airports, schools and offices. This is because of three Sol Robotics proprietary technologies. The first is a shielded commercial grade UV-C source that has both mechanical and electrical safeguards to prevent human exposure. Second, a novel low-cost robotic arm that can effectively deploy the UV-C to a target location. Third, an AI driven sensor fusion algorithm that allows us to detect which objects within a public space are most likely contaminated and in need of disinfection. 

Sol Robotic’s MVP utilizes several novel technologies:

(1) training an AI to detect high-contact surfaces via thermal gradients and LIDAR: In order to disinfect public spaces effectively, it is critical to have the ability to identify surfaces with a high probability of contamination. The team has developed a method for doing this by utilizing AI with LIDAR inputs to identify and tag objects considered to be likely contaminated, such as chairs, tables, and handrails. This information will be paired with data from an IR camera that will be used to collect thermal gradient information from the surroundings in order to inform our AI of which objects have been recently contaminated (e.g. a warm chair). 

(2) Safe deployment of commercial grade UV-C in a public setting: The team is developing a UV-C lamp that is safe for use in public if handled by a robotic system. The device utilizes low-energy radiant UV-C, which is concentrated by a Fresnel lens at a point in close proximity (<2cm) to the shielded lamp covering. By combining this with integrated sensor proximity sensors,  the system can disinfect high-contact surfaces safely in the presence of humans. 

(3) Low-cost robotic arm for deployment of UV-C source: In order to deploy the UV-C source to a target surface, Sol Robotics has developed a low-cost robotic arm with best-in-class reach in order to maximize utility. The arm a new type of linear actuator that is arranged in a parallel structure in order to eliminate torques applied to the system. 

As previously mentioned, There are several new technologies that comprise Sol Robotics’ initial product offering. While these technologies are new, the principles behind each have already been largely proven within academic and commercial spaces. Evidence supporting the tech feasibility is provided below:

1) AI driven sensor fusion algorithms for the detection of high contact surfaces: Object detection using AI and multiple sensor inputs is not a new concept. In fact, there are numerous research groups and companies that have published countless object detection algorithms, such as SSD, RefineDet, and YOLO. Sol Robotics is using these proven methods to develop a hybrid algorithm that best fits our purposes.

2) Deployment of commercial grade UV-C in a public setting: Ultraviolet light has been proven an effective method of killing most pathogens when the correct wavelength and energy intensity is applied. This is supported by the findings of numerous peer-reviewed studies, some of which are listed here: 

[1] https://pubmed.ncbi.nlm.nih.go...

[2] https://bmcinfectdis.biomedcen...

[3] https://www-sciencedirect-com.ezproxyberklee.flo.org/...

These studies have helped Sol Robotics in the design of the UV-C source that will be used in our initial product offering. Our approach for making the product safe, while novel, relies on the use of a simple integrated feedback network that utilizes proximity sensors and mechanical shielding to safeguard those nearby.

3) Low-cost robotic arm for deployment of UV-C source: The team has de-risked this technology with a proof-of-concept (PoC) robotic arm that meets our design requirements. This PoC is highlighted in the product demo link.

The problem that we seek to address is the spread of infectious pathogens through high-contact surfaces in public areas that experience persistently high traffic. The CDC identifies surface contact as a primary source and transmission mechanism for harmful pathogens. A critical aspect of this problem is the presence of traffic, which increases transmission likelihood and compounds the difficulty of preventative measures such as disinfectants. To address this problem, our company will be providing a technology service that will automatically disinfect high-contact surfaces in these locations. The immediate outcomes of this service will be cleaner public spaces, reduced presence of pathogens on surfaces, and a reduced cleaning workload. By reducing the presence of pathogens, human occupants will have a drastically decreased probability of infection, which will, in turn, contribute to improving general public health and increasing resistance to the proliferation of infectious diseases.

Currently, our solution is in the prototype stage of development and is not in service. In one year, we expect to deploy our solution to at least one major local hospital, allowing it to serve approximately 10,000 individuals based on current hospitalization projections. In five years, we seek to increase hospital presence to approximately 100 hospitals and extend our solution to adjacent public venues such as hotels and public transportation centers, which would translate to meaningful service to tens of millions of individuals.

Our company’s short term goal is to deploy our solution to a local hospital and demonstrate a statistically significant decrease in the number of HCAIs over a one-year period of use. Having demonstrated the utility of our technology in a hospital environment, we will seek to partner with multiple healthcare organizations, hospitality companies, and more to expand our technology into as many as 100 public locales internationally. In addition to an expansion into the aforementioned markets, we plan to extend into public transportation and airports with the goal of being implemented in a state-level public transportation infrastructure within five years.

The primary barrier to accomplishing our initial goal of hospital deployment is approval from a candidate hospital’s board of trustees. In order to deploy our solution, we will be required to demonstrate both the practicality and, most importantly, the safety of the system for public use in a hospital setting. In five years, our largest barrier will be to traverse state and federal government regulations surrounding deployment of automated technology in other public venues, such as public transportation centers.

To address the barriers of obtaining private and public approval of our technology, we will leverage resources and standard regulations from governing organizations in the fields of health and robotics such as the CDC, WHO, and NRI. As an example, the NRI has initiated programs designed to accommodate the development of an infrastructure and accompanying set of standards for the proliferation of collaborative robotics. By researching and anticipating the appropriate protocols for human safety surrounding our technology (e.g. UV exposure levels, automation safety factors), we can develop a solution that will satisfy private and public regulations, facilitating our quick transition from development to deployment.

As of today, we have two full-time workers (Justin Hunt and Varun Nalam) and two part-time workers (Andrew Barkan and Zahi Kakish). In addition, we are working with a lawyer, Michael Bane of Wilmer-Hale, who is graciously helping us setting up our C-Corp along with other services toward the development of our startup. Lisa Coca and Dror Sharon of the Intel Corporation are fulfilling roles as expert entrepreneurial advisors to aid in coming up with ideas, attaining future capital investment, and providing advice in navigating startup logistics.

Sol Robotics consists of four roboticists with over 38 years of experience in the field of robotics, including expertise in hardware, software, embedded systems, controls, and AI. We have all graduated from, or are nearing graduation of our PhD programs. The company was formed out of the COVID-19 pandemic as we actively looked for ways to help. Our educations/qualifications are as follows:

• Justin Hunt, PhD Candidate (def. 6/29/2020) in Mechanical Engineering (Robotics), Arizona State University: www.linkedin.com/in/justin-hunt-151530171 

• Varun Nalam, PhD in Mechanical Engineering (Robotics), Arizona State University: https://www.linkedin.com/in/varun-nalam-42969124/

• Zahi Kakish, PhD Candidate in Mechanical Engineering (Robotics), Arizona State University: https://www.linkedin.com/in/zahi-kakish-0b6a6694/

• Andrew Barkan, PhD Candidate in Mechanical Engineering (Robotics), University of California, Berkeley: https://www.linkedin.com/in/andrew-barkan-b1bb1191/

Intel is partnered with us in way of $250,000, 10% equity investment and continuing advisory support.

Our main value proposition is providing disinfection solutions in areas with high human traffic, especially in vulnerable areas such as public places in hospitals, airports, and transit stations using a robotic system. Our business model would be to provide robots as a service (RaaS) to these businesses,who would then be charged a monthly subscription based on the robots used. By using UV-C to disinfect contact surfaces, the businesses can guarantee the best disinfection without any hindrance to their customers and with minimal additional resources or costs. The utilization of these services would ensure that these businesses could serve their customers without endangering their health.

Our main source of revenue would be the monthly subscription service for the robots that can disinfect the high contact surfaces. We would be using a combination of SBIR and raised investment capital for initial research and development along with building the initial fleet for offering Robotics as a Service. The expansion of the fleet will be sponsored by the income from our RaaS model and further investments depending on the demand and expansion plans.

The product that Sol Robotics is looking to bring to the market was born out of the teams’ desire to help with the COVID-19 crisis. After looking for supportive organizations with similar goals, we felt that MIT Solve was the best fit for us given the stage that we are at. We are hoping that Solve can help us overcome product regulatory barriers set by organizations like the FDA and NRI. 

As a small team, we are hoping to get feedback and advice on our product distribution and funding model as we scale. We are also looking for contacts in the manufacturing space so that we can get help moving quickly into production after the MVP is finalized and field tested.

This is somewhat of an open question for our team as we are not sure which organizations are under the umbrella of Solve. Faculty with expertise in AI and Optics to help evaluate and improve our solution would be of great assistance to the team.

The implications of automated disinfection are incredibly significant, and our proposed solution represents a critical step toward the systematic improvement of modern society’s health and its resilience to infectious diseases. In order to make this possible, our solution will rely heavily on the accuracy and robustness of our AI contact tracing software. The detection algorithms we hope to implement will require sensor fusion with a combination of infrared, laser, and visible light imaging data in order to construct an accurate representation of potential high-contact surfaces. Furthermore, the accompanying challenge of navigating a human-occupied space in order to access said surfaces will necessitate novel planning strategies. Integration of our AI with cloud infrastructure will allow us to continuously iterate and improve on our algorithms through a constant stream of critical usage data from the deployment of our systems. Our team is well-posed and uniquely qualified to attack these problems, given our extensive experience in the fields of robotics, deep learning, algorithmic human-robot interaction, and control systems. We plan to use funding from the AI for Humanity Prize to invest in sophisticated sensing units (e.g. infrared and stereoscopic cameras, LIDAR, etc.) to improve our data quality as well as server solutions for training our AI technologies.

Thank you for your consideration.

Sol Robotics is a company with a mission to fight the spread of pathogens globally. Our technology for autonomous robotic disinfection of public spaces has widespread application within hundreds of industries, including hospitals, schools, airports, offices, grocery stores, and sporting arenas.  With so much potential, we require venture capital investment to help finish the development of our MVP and scale. With the right equity partners, Sol Robotics has the potential to deliver an international fleet of robotics systems that could improve the public safety of millions, if not eventually billions of people. We believe that our team is well-positioned to deliver this product to the market, given our extensive experience in robotics, our business acumen, which was honed within the Intel Incubator Program, and most importantly our drive to help in the collective fight against global pathogens. 

Thank you for your consideration.