Lessons learned processes have been deployed in commercial, government, and military organizations since the late 1980s to capture, store, disseminate, and share experiential working knowledge. However, recent studies have shown that software systems for supporting lesson dissemination do not effectively promote knowledge sharing. We found that the problems with these systems are related to their textual representation for lessons and that they are not incorporated into the processes they are intended to support. In this article, we survey lessons learned processes and systems, detail their capabilities and limitations, examine lessons learned system design issues, and identify how artificial intelligence technologies can contribute to knowledge management solutions for these systems.

http://josquin.cs.depaul.edu/~rburke/courses/s04/csc594/docs/weber-etal-01.pdf

Intelligent lessons learned systems

We developed an integrative perspective on the empirical evidence supporting the influence of particular variables on the warning process based on a broad review of the warning literature. The warning process is described in terms of the following four components: notice, encode, comprehend, and comply. Relevant variables are classified as person variables (characteristics of the individual interacting with the warning) and warning variables (characteristics of the warning itself or the context in which the warning appears). This integrative perspective yields general principles about the variables that influence the warning process and serves as a resource for warning developers and as a guide to facilitate effective analysis of warnings. We also identify aspects of the warning process that are not well understood, directions for effective methods of intervention, and a research agenda for future efforts. Actual or potential applications of this research include improving the design of warnings.

Warning research: an integrative perspective.

The purpose of warnings is twofold. The first goal is to inform people so they appreciate potential hazards. The second goal is to change behavior, that is, to redi rect people away from performing unsafe acts that they might other wise perform. With today's tech nology, warnings have become in creasingly necessary. Products, equipment, tools, and the environ ment have become more complex; how they work, their composition, and their inherent hazards are fre

WARNING! Sign and Label Effectiveness:

The development of automated and coordinated driving systems (platooning) is an hot topic today for vehicles and it represents a challenging scenario that heavily relies on distributed control in the presence of wireless communication network. To actuate platooning in a safe way it is necessary to design controllers able to effectively operate on informations exchanged via Inter-Vehicular Communication (IVC) systems despite the presence of unavoidable communication impairments, such as multiple time-varying delays that affect communication links. To this aim in this paper we propose a novel distributed adaptive collaborative control strategy that exploits information coming from connected vehicles to achieve leader synchronization and we analytically demonstrate its stability with a Lyapunov-Krasovskii approach. The effectiveness of the proposed strategy is shown via numerical simulations in P lexe , a state of the art IVC and mobility simulator that includes basic building blocks for platooning.

Adaptive multi-agents synchronization for collaborative driving of autonomous vehicles with multiple communication delays

Red light running is estimated to account for 900 intersection crash fatalities annually. Two principal methods used to reduce red light running involve lengthening the duration of yellow change intervals and automated red light enforcement. The present study evaluated the incremental effects on red light running of first lengthening yellow signal timing, followed by the introduction of red light cameras. At six approaches to two intersections in Philadelphia, Pennsylvania, yellow change intervals were increased by about 1 s, followed several months later by red light camera enforcement. The number of red light violations was monitored before changes were implemented, several weeks after yellow timing changes were made, and about 1 year after commencement of red light camera enforcement. Similar observations were conducted at three comparison intersections in a neighboring state where red light cameras were not used and yellow timing remained constant. Results showed that yellow timing changes reduced red light violations by 36%. The addition of red light camera enforcement further reduced red light violations by 96% beyond levels achieved by the longer yellow timing. This study shows that the provision of adequate yellow signal timing reduces red light running, but longer yellow timing alone does not eliminate the need for better enforcement, which can be provided effectively by red light cameras.

Reducing Red Light Running Through Longer Yellow Signal Timingand Red Light Camera Enforcement: Results of a Field Investigation

The preferred time-headway of drivers in highway conditions is related to the likelihood of rear-end collisions. We studied traffic data from a section of southbound highway 101- a heavily commuted eight-lane freeway between San Francisco and the Silicon valley in California. We observed two parameters that drivers regulate during free flow, rush hour, and heavy traffic conditions: (1) the speed of their vehicle; and (2) the time-headway to the preceding vehicle. During free flow traffic, the preferred speeds show low variation within lanes, but large variations from lane to lane. During rush hour traffic, the time-headway between vehicles varies between 1 and 2 s for a range of traffic speeds. For all traffic conditions a lower limit of 1s is seen in time-headway, even when traffic volume does not push drivers toward tight spacing. The lower limit of 1s is consistent with what was found in several previous studies, but is significantly shorter than the 3s headway that is recommended by driving manuals. The short time-headways observed are within the limit of typical reaction time for braking by alert drivers, but probably lead to occasional accidents given variability in reaction times, decisions, and vehicle braking capabilities, especially when preview information is not available.

Preferred time-headway of highway drivers

The term warning is applied to a variety of stimuli. From a safety standpoint, the most appropriate definition of warning ties it to any information that has the potential to change behavior and prevent accidents. The results of an extensive literature review suggest that warnings are unlikely to be effective unless a series of conditions are met. The failure of many intended warnings, including most on-product warning labels, to reduce accidents reflects the difficulty of overcoming the problems inherent in their use.

What is a Warning and When Will it Work

A variety of potential hazards can be identified for nearly any consumer product, often more than can be practically or effectively addressed with warning labels. Published standards and guidelines for warnings do not offer a reasonable basis for limiting the number and length of warning labels. This paper proposes criteria for the use and design of warning labels based on effectiveness research, accident data, and product-associated risk.

Thomas J. Ayres

Papers

Preferred time-headway of highway drivers

What is a Warning and When Will it Work

Risk and effectiveness criteria for using on-product warnings

Sidewalk Potential Trip Points: A Method for Characterizing Walkways

Passenger-Side Rear-View Mirrors: Driver Behavior and Safety